Commit
•
fbfd857
1
Parent(s):
1cdad4d
Update README.md
Browse files
README.md
CHANGED
@@ -17,28 +17,28 @@ model-index:
|
|
17 |
metrics:
|
18 |
- name: mIoU
|
19 |
type: mIoU
|
20 |
-
value: 77.
|
21 |
- name: IoU Other
|
22 |
type: IoU
|
23 |
-
value:
|
24 |
- name: IoU Ground
|
25 |
type: IoU
|
26 |
-
value: 91.
|
27 |
- name: IoU Vegetation
|
28 |
type: IoU
|
29 |
-
value: 93.
|
30 |
- name: IoU Building
|
31 |
type: IoU
|
32 |
-
value: 90.
|
33 |
- name: IoU Water
|
34 |
type: IoU
|
35 |
-
value:
|
36 |
- name: IoU Bridge
|
37 |
type: IoU
|
38 |
-
value:
|
39 |
- name: IoU Permanent Structure
|
40 |
type: IoU
|
41 |
-
value:
|
42 |
- task:
|
43 |
type: semantic-segmentation
|
44 |
dataset:
|
@@ -206,23 +206,20 @@ The following table gives the class-wise metrics on the test set:
|
|
206 |
|
207 |
**Class**|**IoU**|**Accuracy**|**Precision**|**Recall**|**F1**
|
208 |
-----|---|--------|---------|------|---
|
209 |
-
**other**|
|
210 |
-
**ground**|91.
|
211 |
-
**vegetation**|93.
|
212 |
-
**building**|90.
|
213 |
-
**water**|
|
214 |
-
**bridge**|
|
215 |
-
**permanent structure**|
|
216 |
-
-----|---|--------|---------|------|---
|
217 |
**Macro Average**|77.5|86.7|88.7|84.2|86.2
|
218 |
|
219 |
|
220 |
-
|
221 |
The following illustration gives the resulting confusion matrix :
|
222 |
* Left : normalised acording to rows: rows sum at 100% and the **recall** is on the diagonal of the matrix
|
223 |
* Right : normalised acording to columns: columns sum at 100% and the **precision** is on the diagonal of the matrix
|
224 |
|
225 |
-
|
226 |
<div style="position: relative; text-align: center;">
|
227 |
<p style="margin: 0;">Normalized Confusion Matrices. (a) Recall, (b) Precision)</p>
|
228 |
<img src="FRACTAL-LidarHD_7cl_randlanet-recall_confusion_matrix.excalidraw.png" alt="Confusion matrices" style="width: 70%; display: block; margin: 0 auto;"/>
|
|
|
17 |
metrics:
|
18 |
- name: mIoU
|
19 |
type: mIoU
|
20 |
+
value: 77.5
|
21 |
- name: IoU Other
|
22 |
type: IoU
|
23 |
+
value: 47.5
|
24 |
- name: IoU Ground
|
25 |
type: IoU
|
26 |
+
value: 91.9
|
27 |
- name: IoU Vegetation
|
28 |
type: IoU
|
29 |
+
value: 93.8
|
30 |
- name: IoU Building
|
31 |
type: IoU
|
32 |
+
value: 90.4
|
33 |
- name: IoU Water
|
34 |
type: IoU
|
35 |
+
value: 90.1
|
36 |
- name: IoU Bridge
|
37 |
type: IoU
|
38 |
+
value: 65.2
|
39 |
- name: IoU Permanent Structure
|
40 |
type: IoU
|
41 |
+
value: 63.5
|
42 |
- task:
|
43 |
type: semantic-segmentation
|
44 |
dataset:
|
|
|
206 |
|
207 |
**Class**|**IoU**|**Accuracy**|**Precision**|**Recall**|**F1**
|
208 |
-----|---|--------|---------|------|---
|
209 |
+
**other**|47.5|54.9|77.8|54.9|64.4
|
210 |
+
**ground**|91.9|97.7|93.8|97.7|95.8
|
211 |
+
**vegetation**|93.8|95.6|98.0|95.6|96.8
|
212 |
+
**building**|90.4|93.7|96.2|93.7|95.0
|
213 |
+
**water**|90.1|92.6|97.1|92.6|94.8
|
214 |
+
**bridge**|65.2|96.1|79.3|78.6|79.0
|
215 |
+
**permanent structure**|63.5|76.6|78.9|76.6|77.7
|
|
|
216 |
**Macro Average**|77.5|86.7|88.7|84.2|86.2
|
217 |
|
218 |
|
|
|
219 |
The following illustration gives the resulting confusion matrix :
|
220 |
* Left : normalised acording to rows: rows sum at 100% and the **recall** is on the diagonal of the matrix
|
221 |
* Right : normalised acording to columns: columns sum at 100% and the **precision** is on the diagonal of the matrix
|
222 |
|
|
|
223 |
<div style="position: relative; text-align: center;">
|
224 |
<p style="margin: 0;">Normalized Confusion Matrices. (a) Recall, (b) Precision)</p>
|
225 |
<img src="FRACTAL-LidarHD_7cl_randlanet-recall_confusion_matrix.excalidraw.png" alt="Confusion matrices" style="width: 70%; display: block; margin: 0 auto;"/>
|