Dataset Viewer
Auto-converted to Parquet Duplicate
sample_id
string
task
string
prompt
string
question
string
choices
list
answer
string
image_ids
list
image_paths
list
images_available
bool
image_availability
list
text_rights_status
string
authored_text_license
string
upstream_terms
string
mb-6b6c615b1756ac7f
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 35 km/h (B) 50 km/h (C) 80 km/h (D) 70 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "35 km/h" }, { "label": "B", "text": "50 km/h" }, { "label": "C", "text": "80 km/h" }, { "label": "D", "text": "70 km/h" } ]
B
[ "figure-1" ]
[ "images/mb-6b6c615b1756ac7f/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-234a07e83a322569
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 40 km/h (B) 35 km/h (C) 50 km/h (D) 70 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "40 km/h" }, { "label": "B", "text": "35 km/h" }, { "label": "C", "text": "50 km/h" }, { "label": "D", "text": "70 km/h" } ]
A
[ "figure-1" ]
[ "images/mb-234a07e83a322569/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-5caa8cfdfdc11ed7
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 50 km/h (B) 60 km/h (C) 90 km/h (D) 80 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "50 km/h" }, { "label": "B", "text": "60 km/h" }, { "label": "C", "text": "90 km/h" }, { "label": "D", "text": "80 km/h" } ]
D
[ "figure-1" ]
[ "images/mb-5caa8cfdfdc11ed7/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-554fa9b6ccb036aa
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 30 km/h (B) 15 km/h (C) 25 km/h (D) 20 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "30 km/h" }, { "label": "B", "text": "15 km/h" }, { "label": "C", "text": "25 km/h" }, { "label": "D", "text": "20 km/h" } ]
A
[ "figure-1" ]
[ "images/mb-554fa9b6ccb036aa/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-f310d0b19bd367a0
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 80 km/h (B) 50 km/h (C) 35 km/h (D) 30 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "80 km/h" }, { "label": "B", "text": "50 km/h" }, { "label": "C", "text": "35 km/h" }, { "label": "D", "text": "30 km/h" } ]
B
[ "figure-1" ]
[ "images/mb-f310d0b19bd367a0/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-80e148a898c31f3c
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 70 km/h (B) 50 km/h (C) 80 km/h (D) 60 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "70 km/h" }, { "label": "B", "text": "50 km/h" }, { "label": "C", "text": "80 km/h" }, { "label": "D", "text": "60 km/h" } ]
D
[ "figure-1" ]
[ "images/mb-80e148a898c31f3c/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-9a912ae024a68c0f
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 30 km/h (B) 35 km/h (C) 25 km/h (D) 50 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "30 km/h" }, { "label": "B", "text": "35 km/h" }, { "label": "C", "text": "25 km/h" }, { "label": "D", "text": "50 km/h" } ]
A
[ "figure-1" ]
[ "images/mb-9a912ae024a68c0f/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-ffeacd400eddc146
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 60 km/h (B) 35 km/h (C) 90 km/h (D) 50 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "60 km/h" }, { "label": "B", "text": "35 km/h" }, { "label": "C", "text": "90 km/h" }, { "label": "D", "text": "50 km/h" } ]
A
[ "figure-1" ]
[ "images/mb-ffeacd400eddc146/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-192a607ca35b9f6b
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 60 km/h (B) 70 km/h (C) 50 km/h (D) 35 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "60 km/h" }, { "label": "B", "text": "70 km/h" }, { "label": "C", "text": "50 km/h" }, { "label": "D", "text": "35 km/h" } ]
C
[ "figure-1" ]
[ "images/mb-192a607ca35b9f6b/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-2efc4b260248ce49
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 90 km/h (B) 40 km/h (C) 60 km/h (D) 50 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "90 km/h" }, { "label": "B", "text": "40 km/h" }, { "label": "C", "text": "60 km/h" }, { "label": "D", "text": "50 km/h" } ]
C
[ "figure-1" ]
[ "images/mb-2efc4b260248ce49/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-6859bd1350256b0f
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 20 km/h (B) 25 km/h (C) 35 km/h (D) 30 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "20 km/h" }, { "label": "B", "text": "25 km/h" }, { "label": "C", "text": "35 km/h" }, { "label": "D", "text": "30 km/h" } ]
D
[ "figure-1" ]
[ "images/mb-6859bd1350256b0f/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-3e4801b80db1799a
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 80 km/h (B) 60 km/h (C) 70 km/h (D) 50 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "80 km/h" }, { "label": "B", "text": "60 km/h" }, { "label": "C", "text": "70 km/h" }, { "label": "D", "text": "50 km/h" } ]
D
[ "figure-1" ]
[ "images/mb-3e4801b80db1799a/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-56213757b13ac0ee
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 60 km/h (B) 70 km/h (C) 40 km/h (D) 50 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "60 km/h" }, { "label": "B", "text": "70 km/h" }, { "label": "C", "text": "40 km/h" }, { "label": "D", "text": "50 km/h" } ]
C
[ "figure-1" ]
[ "images/mb-56213757b13ac0ee/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-5aa0b96d6f93ba45
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 15 km/h (B) 5 km/h (C) 20 km/h (D) 10 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "15 km/h" }, { "label": "B", "text": "5 km/h" }, { "label": "C", "text": "20 km/h" }, { "label": "D", "text": "10 km/h" } ]
B
[ "figure-1" ]
[ "images/mb-5aa0b96d6f93ba45/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-58e82d0165e555a1
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 100 km/h (B) 120 km/h (C) 90 km/h (D) 110 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "100 km/h" }, { "label": "B", "text": "120 km/h" }, { "label": "C", "text": "90 km/h" }, { "label": "D", "text": "110 km/h" } ]
B
[ "figure-1" ]
[ "images/mb-58e82d0165e555a1/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-ba6373d6a86046d2
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 100 km/h (B) 80 km/h (C) 50 km/h (D) 70 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "100 km/h" }, { "label": "B", "text": "80 km/h" }, { "label": "C", "text": "50 km/h" }, { "label": "D", "text": "70 km/h" } ]
B
[ "figure-1" ]
[ "images/mb-ba6373d6a86046d2/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-66756d0965807036
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 25 km/h (B) 40 km/h (C) 30 km/h (D) 15 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "25 km/h" }, { "label": "B", "text": "40 km/h" }, { "label": "C", "text": "30 km/h" }, { "label": "D", "text": "15 km/h" } ]
C
[ "figure-1" ]
[ "images/mb-66756d0965807036/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-446343e7040cb143
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 20 km/h (B) 15 km/h (C) 5 km/h (D) 10 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "20 km/h" }, { "label": "B", "text": "15 km/h" }, { "label": "C", "text": "5 km/h" }, { "label": "D", "text": "10 km/h" } ]
C
[ "figure-1" ]
[ "images/mb-446343e7040cb143/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-406dcc366fbfeb24
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 20 km/h (B) 10 km/h (C) 15 km/h (D) 5 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "20 km/h" }, { "label": "B", "text": "10 km/h" }, { "label": "C", "text": "15 km/h" }, { "label": "D", "text": "5 km/h" } ]
D
[ "figure-1" ]
[ "images/mb-406dcc366fbfeb24/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-6f3b597f0d45b57b
Distant Sign Reading
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show? (A) 70 km/h (B) 60 km/h (C) 50 km/h (D) 80 km/h Answer with only the correct option letter.
This is a 2048x2048 dashcam-style street photo containing exactly one speed-limit sign, which is far away and only a few dozen pixels tall. What speed limit does it show?
[ { "label": "A", "text": "70 km/h" }, { "label": "B", "text": "60 km/h" }, { "label": "C", "text": "50 km/h" }, { "label": "D", "text": "80 km/h" } ]
A
[ "figure-1" ]
[ "images/mb-6f3b597f0d45b57b/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC (upstream does not specify version)
mb-747486ace23e3d18
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.5930, 0.2753, 0.8530, 0.5353]" }, { "label": "B", "text": "[0.2962, 0.7356, 0.5562, 0.9956]" }, { "label": "C", "text": "[0.2568, 0.4544, 0.5168, 0.7144]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-747486ace23e3d18/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-00428bcad2f73883
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.1492, 0.1241, 0.4092, 0.3841]" }, { "label": "B", "text": "[0.2521, 0.7374, 0.5121, 0.9974]" }, { "label": "C", "text": "[0.1590, 0.4567, 0.4190, 0.7167]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-00428bcad2f73883/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-cc045bb0a3268bcc
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.6277, 0.1307, 0.8877, 0.3907]" }, { "label": "B", "text": "[0.3890, 0.5020, 0.6490, 0.7620]" }, { "label": "C", "text": "[0.7171, 0.6182, 0.9771, 0.8782]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-cc045bb0a3268bcc/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-b3061a2046eb8714
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.4762, 0.3370, 0.7362, 0.5970]" }, { "label": "B", "text": "[0.1953, 0.0416, 0.4553, 0.3016]" }, { "label": "C", "text": "[0.6460, 0.0148, 0.9060, 0.2748]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-b3061a2046eb8714/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-bd2c78f872ebf000
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.5997, 0.2296, 0.8597, 0.4896]" }, { "label": "B", "text": "[0.4733, 0.5681, 0.7333, 0.8281]" }, { "label": "C", "text": "[0.0503, 0.0011, 0.3103, 0.2611]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-bd2c78f872ebf000/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-a7dccd74e65fd0cd
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.0656, 0.6717, 0.3256, 0.9317]" }, { "label": "B", "text": "[0.7326, 0.7223, 0.9926, 0.9823]" }, { "label": "C", "text": "[0.2117, 0.3128, 0.4717, 0.5728]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-a7dccd74e65fd0cd/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-4bca1f5d97790d3b
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.1600, 0.2485, 0.4200, 0.5085]" }, { "label": "B", "text": "[0.5023, 0.1040, 0.7623, 0.3640]" }, { "label": "C", "text": "[0.2576, 0.5292, 0.5176, 0.7892]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-4bca1f5d97790d3b/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-d6c054e62497df21
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.6972, 0.4658, 0.9572, 0.7258]" }, { "label": "B", "text": "[0.3963, 0.6151, 0.6563, 0.8751]" }, { "label": "C", "text": "[0.1542, 0.0984, 0.4142, 0.3584]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-d6c054e62497df21/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-8f9f360092350921
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.2232, 0.6574, 0.4832, 0.9174]" }, { "label": "B", "text": "[0.4534, 0.0773, 0.7134, 0.3373]" }, { "label": "C", "text": "[0.0935, 0.1848, 0.3535, 0.4448]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-8f9f360092350921/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-97bb51113e15ca40
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.1617, 0.6831, 0.4217, 0.9431]" }, { "label": "B", "text": "[0.4928, 0.4148, 0.7528, 0.6748]" }, { "label": "C", "text": "[0.0847, 0.3128, 0.3447, 0.5728]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-97bb51113e15ca40/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-af7965e6fd6817b5
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.1905, 0.0225, 0.4505, 0.2825]" }, { "label": "B", "text": "[0.3395, 0.3686, 0.5995, 0.6286]" }, { "label": "C", "text": "[0.0550, 0.3279, 0.3150, 0.5879]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-af7965e6fd6817b5/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-d3a3775f9536cb27
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.3603, 0.2385, 0.6203, 0.4985]" }, { "label": "B", "text": "[0.1670, 0.5279, 0.4270, 0.7879]" }, { "label": "C", "text": "[0.6897, 0.4315, 0.9497, 0.6915]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-d3a3775f9536cb27/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-23a448fe7cd593e1
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.6510, 0.7097, 0.9110, 0.9697]" }, { "label": "B", "text": "[0.3423, 0.6696, 0.6023, 0.9296]" }, { "label": "C", "text": "[0.4707, 0.2335, 0.7307, 0.4935]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-23a448fe7cd593e1/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-4f61b2506a83e3e2
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.2734, 0.2883, 0.5334, 0.5483]" }, { "label": "B", "text": "[0.0432, 0.5805, 0.3032, 0.8405]" }, { "label": "C", "text": "[0.4470, 0.7315, 0.7070, 0.9915]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-4f61b2506a83e3e2/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-1c83dac7c9d22ac7
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.0575, 0.5668, 0.3175, 0.8268]" }, { "label": "B", "text": "[0.5992, 0.0812, 0.8592, 0.3412]" }, { "label": "C", "text": "[0.4355, 0.6462, 0.6955, 0.9062]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-1c83dac7c9d22ac7/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-0aaedb3a8711bd71
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.4455, 0.3454, 0.7055, 0.6054]" }, { "label": "B", "text": "[0.6377, 0.6321, 0.8977, 0.8921]" }, { "label": "C", "text": "[0.0291, 0.0533, 0.2891, 0.3133]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-0aaedb3a8711bd71/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-9df2d3da5373bc89
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.4476, 0.0917, 0.7076, 0.3517]" }, { "label": "B", "text": "[0.4430, 0.5067, 0.7030, 0.7667]" }, { "label": "C", "text": "[0.0133, 0.5507, 0.2733, 0.8107]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-9df2d3da5373bc89/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-30d46731a5c839c3
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.0100, 0.4156, 0.2700, 0.6756]" }, { "label": "B", "text": "[0.3480, 0.3166, 0.6080, 0.5766]" }, { "label": "C", "text": "[0.6544, 0.4412, 0.9144, 0.7012]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-30d46731a5c839c3/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-0afe1e6b9469a546
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.6196, 0.1721, 0.8796, 0.4321]" }, { "label": "B", "text": "[0.6130, 0.6887, 0.8730, 0.9487]" }, { "label": "C", "text": "[0.0653, 0.2646, 0.3253, 0.5246]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-0afe1e6b9469a546/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-bc4e9cdd1da558d5
Pathology Zoom
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
This is a region of an H&E-stained whole-slide image of a lymph node, shown at full scanner resolution. Metastatic tumour tissue can only be told from normal lymph node tissue by nuclear morphology, which is not resolvable unless you magnify. The candidate regions below are normalized [x1, y1, x2, y2] in [0, 1] (origin...
[ { "label": "A", "text": "[0.4139, 0.5420, 0.6739, 0.8020]" }, { "label": "B", "text": "[0.4461, 0.0353, 0.7061, 0.2953]" }, { "label": "C", "text": "[0.1491, 0.1602, 0.4091, 0.4202]" }, { "label": "D", "text": "No metastatic tumour in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-bc4e9cdd1da558d5/figure-1.jpg" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC0-1.0
mb-7051be7689aa82e7
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.8172, 0.3742, 0.9372, 0.4942]" }, { "label": "B", "text": "[0.7366, 0.5509, 0.8566, 0.6709]" }, { "label": "C", "text": "[0.2050, 0.4099, 0.3250, 0.5299]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-7051be7689aa82e7/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-317877c7b2bae170
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.6006, 0.0794, 0.7537, 0.2325]" }, { "label": "B", "text": "[0.3024, 0.2732, 0.4555, 0.4263]" }, { "label": "C", "text": "[0.0876, 0.2950, 0.2407, 0.4481]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-317877c7b2bae170/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-9eca7f14ea6cf650
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.2873, 0.5006, 0.4601, 0.6734]" }, { "label": "B", "text": "[0.5349, 0.4350, 0.7077, 0.6078]" }, { "label": "C", "text": "[0.0505, 0.3827, 0.2233, 0.5555]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-9eca7f14ea6cf650/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-2425870bd8c1f7c7
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.8038, 0.3349, 0.9238, 0.4549]" }, { "label": "B", "text": "[0.4884, 0.5421, 0.6084, 0.6621]" }, { "label": "C", "text": "[0.2979, 0.4122, 0.4179, 0.5322]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-2425870bd8c1f7c7/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-6cd994bc9dbfe0c6
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.2813, 0.4458, 0.4205, 0.5850]" }, { "label": "B", "text": "[0.1022, 0.3406, 0.2414, 0.4798]" }, { "label": "C", "text": "[0.7155, 0.5437, 0.8547, 0.6829]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-6cd994bc9dbfe0c6/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-42a5f40c0176c05e
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.8021, 0.5258, 0.9221, 0.6458]" }, { "label": "B", "text": "[0.2847, 0.4143, 0.4047, 0.5343]" }, { "label": "C", "text": "[0.0092, 0.1303, 0.1292, 0.2503]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-42a5f40c0176c05e/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-41189943b9b44ac5
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.8514, 0.4218, 0.9730, 0.5434]" }, { "label": "B", "text": "[0.0646, 0.4794, 0.1862, 0.6010]" }, { "label": "C", "text": "[0.2666, 0.4479, 0.3882, 0.5695]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-41189943b9b44ac5/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-2ab480b5a502c87d
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.3740, 0.4703, 0.4940, 0.5903]" }, { "label": "B", "text": "[0.2125, 0.3591, 0.3325, 0.4791]" }, { "label": "C", "text": "[0.8358, 0.4179, 0.9558, 0.5379]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-2ab480b5a502c87d/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-593f353366f051e1
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.4576, 0.3913, 0.5872, 0.5209]" }, { "label": "B", "text": "[0.1680, 0.5535, 0.2976, 0.6831]" }, { "label": "C", "text": "[0.7110, 0.4526, 0.8406, 0.5822]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-593f353366f051e1/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-333003f715ad05ca
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.5229, 0.5108, 0.6733, 0.6612]" }, { "label": "B", "text": "[0.7790, 0.5217, 0.9294, 0.6721]" }, { "label": "C", "text": "[0.7594, 0.3316, 0.9098, 0.4820]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-333003f715ad05ca/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-57fafdbbe8d3720c
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.7171, 0.4124, 0.8371, 0.5324]" }, { "label": "B", "text": "[0.2575, 0.5188, 0.3775, 0.6388]" }, { "label": "C", "text": "[0.0465, 0.4524, 0.1665, 0.5724]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-57fafdbbe8d3720c/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-f59203f364850848
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.0048, 0.3206, 0.1948, 0.5106]" }, { "label": "B", "text": "[0.3519, 0.5034, 0.5419, 0.6934]" }, { "label": "C", "text": "[0.3999, 0.2038, 0.5899, 0.3938]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-f59203f364850848/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-622c17de81e60866
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.6719, 0.5158, 0.8047, 0.6486]" }, { "label": "B", "text": "[0.3764, 0.5265, 0.5092, 0.6593]" }, { "label": "C", "text": "[0.2224, 0.7852, 0.3552, 0.9180]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
B
[ "figure-1" ]
[ "images/mb-622c17de81e60866/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-adbbcde489fb2aa0
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.3628, 0.3981, 0.5222, 0.5575]" }, { "label": "B", "text": "[0.6999, 0.5505, 0.8593, 0.7099]" }, { "label": "C", "text": "[0.0644, 0.4076, 0.2238, 0.5670]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
A
[ "figure-1" ]
[ "images/mb-adbbcde489fb2aa0/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-7ef40cd8ed693f74
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.7793, 0.1017, 0.9777, 0.3001]" }, { "label": "B", "text": "[0.7227, 0.3287, 0.9211, 0.5271]" }, { "label": "C", "text": "[0.3310, 0.1979, 0.5294, 0.3963]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
C
[ "figure-1" ]
[ "images/mb-7ef40cd8ed693f74/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-c63a66fc4dab3660
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.3544, 0.3395, 0.4976, 0.4827]" }, { "label": "B", "text": "[0.6391, 0.6743, 0.7823, 0.8175]" }, { "label": "C", "text": "[0.2620, 0.1225, 0.4052, 0.2657]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-c63a66fc4dab3660/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-e9f122484a27b8ea
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.2332, 0.3336, 0.4232, 0.5236]" }, { "label": "B", "text": "[0.5929, 0.1696, 0.7829, 0.3596]" }, { "label": "C", "text": "[0.1891, 0.5732, 0.3791, 0.7632]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-e9f122484a27b8ea/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-27ad357be8d2130e
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.4014, 0.7384, 0.5246, 0.8616]" }, { "label": "B", "text": "[0.4151, 0.5645, 0.5383, 0.6877]" }, { "label": "C", "text": "[0.6484, 0.1796, 0.7716, 0.3028]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-27ad357be8d2130e/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-6f6a496c00edf4f8
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.3032, 0.4388, 0.4935, 0.6291]" }, { "label": "B", "text": "[0.4064, 0.6903, 0.5967, 0.8806]" }, { "label": "C", "text": "[0.7887, 0.3082, 0.9790, 0.4985]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-6f6a496c00edf4f8/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-f8dc3ed50b90c619
Micro Scratch Detection
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect? (...
The image is a high-resolution photo of an industrial part surface. The candidate regions below are given as normalized [x1, y1, x2, y2] with values in [0, 1] (origin at the top-left corner). At most one of them contains a small surface defect such as a scratch, cut, crack or thread. Which region contains the defect?
[ { "label": "A", "text": "[0.7212, 0.4496, 0.8896, 0.6180]" }, { "label": "B", "text": "[0.3256, 0.6438, 0.4940, 0.8122]" }, { "label": "C", "text": "[0.5265, 0.2719, 0.6949, 0.4403]" }, { "label": "D", "text": "No defect in any of the above regions." } ]
D
[ "figure-1" ]
[ "images/mb-f8dc3ed50b90c619/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-NC-SA-4.0
mb-ac3a6f24384272f1
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 58 (B) 46 (C) 65 (D) 70 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "58" }, { "label": "B", "text": "46" }, { "label": "C", "text": "65" }, { "label": "D", "text": "70" } ]
A
[ "figure-1" ]
[ "images/mb-ac3a6f24384272f1/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-f71dd7180c4f8237
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 65 (B) 61 (C) 56 (D) 52 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "65" }, { "label": "B", "text": "61" }, { "label": "C", "text": "56" }, { "label": "D", "text": "52" } ]
D
[ "figure-1" ]
[ "images/mb-f71dd7180c4f8237/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-e5734a979751c230
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 46 (B) 50 (C) 55 (D) 59 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "46" }, { "label": "B", "text": "50" }, { "label": "C", "text": "55" }, { "label": "D", "text": "59" } ]
A
[ "figure-1" ]
[ "images/mb-e5734a979751c230/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-c1d3b79b54a3246c
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 32 (B) 46 (C) 41 (D) 36 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "32" }, { "label": "B", "text": "46" }, { "label": "C", "text": "41" }, { "label": "D", "text": "36" } ]
B
[ "figure-1" ]
[ "images/mb-c1d3b79b54a3246c/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-f188d33f297b1956
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 37 (B) 46 (C) 56 (D) 32 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "37" }, { "label": "B", "text": "46" }, { "label": "C", "text": "56" }, { "label": "D", "text": "32" } ]
B
[ "figure-1" ]
[ "images/mb-f188d33f297b1956/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-bb21473c1ff8033e
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 70 (B) 60 (C) 65 (D) 56 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "70" }, { "label": "B", "text": "60" }, { "label": "C", "text": "65" }, { "label": "D", "text": "56" } ]
D
[ "figure-1" ]
[ "images/mb-bb21473c1ff8033e/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-810ee8b895d10dd5
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 43 (B) 39 (C) 51 (D) 56 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "43" }, { "label": "B", "text": "39" }, { "label": "C", "text": "51" }, { "label": "D", "text": "56" } ]
C
[ "figure-1" ]
[ "images/mb-810ee8b895d10dd5/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-0ae9e378788c366e
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 47 (B) 67 (C) 59 (D) 72 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "47" }, { "label": "B", "text": "67" }, { "label": "C", "text": "59" }, { "label": "D", "text": "72" } ]
C
[ "figure-1" ]
[ "images/mb-0ae9e378788c366e/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-28f8e94eb505fa7f
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 49 (B) 58 (C) 62 (D) 53 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "49" }, { "label": "B", "text": "58" }, { "label": "C", "text": "62" }, { "label": "D", "text": "53" } ]
A
[ "figure-1" ]
[ "images/mb-28f8e94eb505fa7f/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-a3a97321e0531584
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 34 (B) 38 (C) 30 (D) 44 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "34" }, { "label": "B", "text": "38" }, { "label": "C", "text": "30" }, { "label": "D", "text": "44" } ]
D
[ "figure-1" ]
[ "images/mb-a3a97321e0531584/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-e8dbf59bfee0d900
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 65 (B) 52 (C) 61 (D) 57 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "65" }, { "label": "B", "text": "52" }, { "label": "C", "text": "61" }, { "label": "D", "text": "57" } ]
A
[ "figure-1" ]
[ "images/mb-e8dbf59bfee0d900/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-9c6d98c679b553b6
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 73 (B) 67 (C) 63 (D) 59 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "73" }, { "label": "B", "text": "67" }, { "label": "C", "text": "63" }, { "label": "D", "text": "59" } ]
D
[ "figure-1" ]
[ "images/mb-9c6d98c679b553b6/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-8ec4543aeb64173d
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 33 (B) 29 (C) 42 (D) 37 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "33" }, { "label": "B", "text": "29" }, { "label": "C", "text": "42" }, { "label": "D", "text": "37" } ]
C
[ "figure-1" ]
[ "images/mb-8ec4543aeb64173d/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-3e7de3fad9fbd565
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 57 (B) 61 (C) 70 (D) 66 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "57" }, { "label": "B", "text": "61" }, { "label": "C", "text": "70" }, { "label": "D", "text": "66" } ]
C
[ "figure-1" ]
[ "images/mb-3e7de3fad9fbd565/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-c33e6f674a4a7152
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 67 (B) 53 (C) 62 (D) 58 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "67" }, { "label": "B", "text": "53" }, { "label": "C", "text": "62" }, { "label": "D", "text": "58" } ]
B
[ "figure-1" ]
[ "images/mb-c33e6f674a4a7152/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-8446c09a91c40c8f
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 48 (B) 56 (C) 52 (D) 44 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "48" }, { "label": "B", "text": "56" }, { "label": "C", "text": "52" }, { "label": "D", "text": "44" } ]
D
[ "figure-1" ]
[ "images/mb-8446c09a91c40c8f/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-de8b89effe027b72
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 32 (B) 41 (C) 36 (D) 27 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "32" }, { "label": "B", "text": "41" }, { "label": "C", "text": "36" }, { "label": "D", "text": "27" } ]
B
[ "figure-1" ]
[ "images/mb-de8b89effe027b72/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-694decbd4995ae16
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 44 (B) 35 (C) 48 (D) 40 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "44" }, { "label": "B", "text": "35" }, { "label": "C", "text": "48" }, { "label": "D", "text": "40" } ]
C
[ "figure-1" ]
[ "images/mb-694decbd4995ae16/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-f85cbd3902c4ecc9
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 56 (B) 68 (C) 46 (D) 52 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "56" }, { "label": "B", "text": "68" }, { "label": "C", "text": "46" }, { "label": "D", "text": "52" } ]
A
[ "figure-1" ]
[ "images/mb-f85cbd3902c4ecc9/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-012ba7748d1e6cf8
Dense Small Object Counting
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total? (A) 27 (B) 41 (C) 37 (D) 33 Answer with only the correct option letter.
The figure is an aerial photo of a parking lot. Count every car visible in the image, including partially visible ones. How many cars are there in total?
[ { "label": "A", "text": "27" }, { "label": "B", "text": "41" }, { "label": "C", "text": "37" }, { "label": "D", "text": "33" } ]
B
[ "figure-1" ]
[ "images/mb-012ba7748d1e6cf8/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
CC-BY-4.0
mb-df28a90356748643
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 333 (B) 92 (C) 272 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "333" }, { "label": "B", "text": "92" }, { "label": "C", "text": "272" }, { "label": "D", "text": "None of the above restores it." } ]
B
[ "figure-1" ]
[ "images/mb-df28a90356748643/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-adf34c68ca59e5b7
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 299 (B) 119 (C) 196 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "299" }, { "label": "B", "text": "119" }, { "label": "C", "text": "196" }, { "label": "D", "text": "None of the above restores it." } ]
B
[ "figure-1" ]
[ "images/mb-adf34c68ca59e5b7/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-02e54780c0c7bb6d
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 149 (B) 359 (C) 179 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "149" }, { "label": "B", "text": "359" }, { "label": "C", "text": "179" }, { "label": "D", "text": "None of the above restores it." } ]
C
[ "figure-1" ]
[ "images/mb-02e54780c0c7bb6d/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-318c9fff004b1318
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 53 (B) 202 (C) 233 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "53" }, { "label": "B", "text": "202" }, { "label": "C", "text": "233" }, { "label": "D", "text": "None of the above restores it." } ]
D
[ "figure-1" ]
[ "images/mb-318c9fff004b1318/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-8c6feb52dd203b72
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 152 (B) 121 (C) 332 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "152" }, { "label": "B", "text": "121" }, { "label": "C", "text": "332" }, { "label": "D", "text": "None of the above restores it." } ]
C
[ "figure-1" ]
[ "images/mb-8c6feb52dd203b72/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-4529068c036c31d0
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 193 (B) 234 (C) 13 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "193" }, { "label": "B", "text": "234" }, { "label": "C", "text": "13" }, { "label": "D", "text": "None of the above restores it." } ]
D
[ "figure-1" ]
[ "images/mb-4529068c036c31d0/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-bbaafb8777d5765e
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 208 (B) 291 (C) 111 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "208" }, { "label": "B", "text": "291" }, { "label": "C", "text": "111" }, { "label": "D", "text": "None of the above restores it." } ]
C
[ "figure-1" ]
[ "images/mb-bbaafb8777d5765e/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-05c8a42db094bf38
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 195 (B) 284 (C) 15 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "195" }, { "label": "B", "text": "284" }, { "label": "C", "text": "15" }, { "label": "D", "text": "None of the above restores it." } ]
A
[ "figure-1" ]
[ "images/mb-05c8a42db094bf38/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-bd813cffff6dcfcf
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 232 (B) 286 (C) 52 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "232" }, { "label": "B", "text": "286" }, { "label": "C", "text": "52" }, { "label": "D", "text": "None of the above restores it." } ]
C
[ "figure-1" ]
[ "images/mb-bd813cffff6dcfcf/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-bba2cd753cf7091b
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 272 (B) 92 (C) 186 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "272" }, { "label": "B", "text": "92" }, { "label": "C", "text": "186" }, { "label": "D", "text": "None of the above restores it." } ]
B
[ "figure-1" ]
[ "images/mb-bba2cd753cf7091b/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-6df6bf80b4d84c3a
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 303 (B) 123 (C) 150 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "303" }, { "label": "B", "text": "123" }, { "label": "C", "text": "150" }, { "label": "D", "text": "None of the above restores it." } ]
A
[ "figure-1" ]
[ "images/mb-6df6bf80b4d84c3a/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-b0248ecb205865c0
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 276 (B) 96 (C) 308 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "276" }, { "label": "B", "text": "96" }, { "label": "C", "text": "308" }, { "label": "D", "text": "None of the above restores it." } ]
B
[ "figure-1" ]
[ "images/mb-b0248ecb205865c0/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-8a3bc37c3ee16c9b
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 66 (B) 184 (C) 4 (D) None of the above restores it. Answer with o...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "66" }, { "label": "B", "text": "184" }, { "label": "C", "text": "4" }, { "label": "D", "text": "None of the above restores it." } ]
D
[ "figure-1" ]
[ "images/mb-8a3bc37c3ee16c9b/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-72b585521cd6ad8a
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 339 (B) 32 (C) 212 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "339" }, { "label": "B", "text": "32" }, { "label": "C", "text": "212" }, { "label": "D", "text": "None of the above restores it." } ]
A
[ "figure-1" ]
[ "images/mb-72b585521cd6ad8a/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-d00a380c592984c5
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 206 (B) 327 (C) 147 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "206" }, { "label": "B", "text": "327" }, { "label": "C", "text": "147" }, { "label": "D", "text": "None of the above restores it." } ]
A
[ "figure-1" ]
[ "images/mb-d00a380c592984c5/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-c5802be4993860bb
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 325 (B) 176 (C) 145 (D) None of the above restores it. Answer wit...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "325" }, { "label": "B", "text": "176" }, { "label": "C", "text": "145" }, { "label": "D", "text": "None of the above restores it." } ]
D
[ "figure-1" ]
[ "images/mb-c5802be4993860bb/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-53abf50ecc1a70b1
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 236 (B) 266 (C) 56 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "236" }, { "label": "B", "text": "266" }, { "label": "C", "text": "56" }, { "label": "D", "text": "None of the above restores it." } ]
C
[ "figure-1" ]
[ "images/mb-53abf50ecc1a70b1/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-e6c5abfe0e4fba57
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 314 (B) 70 (C) 250 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "314" }, { "label": "B", "text": "70" }, { "label": "C", "text": "250" }, { "label": "D", "text": "None of the above restores it." } ]
A
[ "figure-1" ]
[ "images/mb-e6c5abfe0e4fba57/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-89cf48df23b0c5cd
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 277 (B) 195 (C) 97 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "277" }, { "label": "B", "text": "195" }, { "label": "C", "text": "97" }, { "label": "D", "text": "None of the above restores it." } ]
B
[ "figure-1" ]
[ "images/mb-89cf48df23b0c5cd/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
mb-ee6f8bfc2acb8c26
Rotated Text OCR
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation? (A) 245 (B) 292 (C) 65 (D) None of the above restores it. Answer with...
The figure is a scanned document page that has been rotated counterclockwise by an unknown angle. Rotating the figure further counterclockwise by which of the following angles (in degrees) restores the page to its normal upright reading orientation?
[ { "label": "A", "text": "245" }, { "label": "B", "text": "292" }, { "label": "C", "text": "65" }, { "label": "D", "text": "None of the above restores it." } ]
D
[ "figure-1" ]
[ "images/mb-ee6f8bfc2acb8c26/figure-1.png" ]
true
[ "bundled" ]
upstream_terms_recorded
CC-BY-4.0
No upstream pixel or annotation content exported in the text fields; source image rights remain separate.
End of preview. Expand in Data Studio

MosaicBench

560 questions across 28 task families, with 20 questions per family, in the original test split. Original prompts, ordered candidate answers and correct option labels are retained. No samples are replaced or silently dropped.

Images and references

780 processed images are included. All 100 COCO-derived figures are reference-only: the manifest records original filenames, COCO IDs, original-image URLs and Flickr photo pages. These 100 figure references correspond to 20 original COCO images. 540 questions have all required images in the package. The other 20 retain their full text and answers. Fetching an original COCO image alone does not recreate the cropped/rotated puzzle figures; no automatic reconstruction is claimed.

image_manifest.jsonl records processed-figure SHA-256 checksums, ordered image references, source terms and availability. These checksums refer to the benchmark figures, not the downloadable original photos. images_available describes the package contents, not later local additions. No benchmark scores are included.

Load locally

from pathlib import Path
from datasets import load_dataset
from PIL import Image
root = Path('.')
dataset = load_dataset('json', data_files={'test': str(root / 'data/test.jsonl')})['test']
runnable = dataset.filter(lambda row: row['images_available'])
row = runnable[0]
images = [Image.open(root / p).convert('RGB') for p in row['image_paths']]

Compare predicted option labels with answer; report evaluated sample IDs and task coverage. A 540-question run is not the complete 560-question benchmark. The rotation question follows the forward counterclockwise relation stated in its prompt. Do not attribute scores from changed question definitions without rerunning evaluation.

For legitimately obtained exact processed figures, restore_images.py can match local files by checksum. It neither downloads originals nor generates derivatives.

Sources and terms

Read LICENSE.md, ATTRIBUTION.md, BDD100K-LICENSE.txt and the per-image manifest. Terms vary by source; some have noncommercial/share-alike requirements, and some lack established redistribution permission. rights_verification and text_rights_status preserve that distinction. Source attribution does not replace permission. No blanket third-party rights grant is made here.

Project links

Load from Hugging Face

from datasets import load_dataset
from huggingface_hub import snapshot_download
from pathlib import Path
from PIL import Image

data = load_dataset("gengyuanmax/MosaicBench", split="test")
root = Path(snapshot_download("gengyuanmax/MosaicBench", repo_type="dataset"))
row = next(row for row in data if row["images_available"])
images = [Image.open(root / p).convert("RGB") for p in row["image_paths"]]

The table contains relative image paths. The local image-loading example resolves them from the downloaded snapshot; it does not rely on automatic viewer decoding.

Downloads last month
-

Paper for gengyuanmax/MosaicBench