Datasets:

Gen3DF
/

tennacs_ply

text stringlengths 16 177
colorHeight = 968
colorToDepthExtrinsics = 0.999263 -0.010031 0.037048 -0.038549 0.009933 0.999947 0.002817 -0.004287 -0.037074 -0.002447 0.999310 -0.020574 -0.000000 0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 570.924255
fy_color = 1170.187988
fy_depth = 570.924316
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 784
numDepthFrames = 784
numIMUmeasurements = 1632
colorHeight = 968
colorToDepthExtrinsics = 0.999263 -0.010031 0.037048 -0.038549 0.009933 0.999947 0.002817 -0.004287 -0.037074 -0.002447 0.999310 -0.020574 -0.000000 0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 570.924255
fy_color = 1170.187988
fy_depth = 570.924316
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 1183
numDepthFrames = 1183
numIMUmeasurements = 2460
colorHeight = 968
colorToDepthExtrinsics = 0.999263 -0.010031 0.037048 -0.038549 0.009933 0.999947 0.002817 -0.004287 -0.037074 -0.002447 0.999310 -0.020574 -0.000000 0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 570.924255
fy_color = 1170.187988
fy_depth = 570.924316
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 936
numDepthFrames = 936
numIMUmeasurements = 1951
appVersionId = Scanner.v1.20161011132740
colorHeight = 968
colorToDepthExtrinsics = 0.999967 0.005619 -0.005838 -0.037692 -0.005643 0.999975 -0.004188 -0.003544 0.005814 0.004221 0.999974 -0.022235 -0.000000 -0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 570.707642
fy_color = 1170.187988
fy_depth = 570.707642
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 1807
numDepthFrames = 1807
numIMUmeasurements = 3615
appVersionId = Scanner.v1.20161020190252
colorHeight = 968
colorToDepthExtrinsics = 0.999999 0.001589 -0.000662 -0.037792 -0.001590 0.999999 -0.000655 -0.003775 0.000661 0.000656 1.000000 -0.022027 -0.000000 -0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 570.727112
fy_color = 1170.187988
fy_depth = 570.727173
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 3361
numDepthFrames = 3361
numIMUmeasurements = 6999
appVersionId = Scanner.v1.20161026133258
colorHeight = 968
colorToDepthExtrinsics = 1.000000 -0.000000 0.000000 -0.037800 -0.000000 1.000000 -0.000000 -0.003850 0.000000 -0.000000 1.000000 -0.022000 -0.000000 0.000000 -0.000000 1.000000
colorWidth = 1296
depthHeight = 480
depthWidth = 640
fx_color = 1170.187988
fx_depth = 571.869507
fy_color = 1170.187988
fy_depth = 571.869507
mx_color = 647.750000
mx_depth = 319.500000
my_color = 483.750000
my_depth = 239.500000
numColorFrames = 4787
numDepthFrames = 4787
numIMUmeasurements = 9821
appVersionId = Scanner.v1.20161011132740

End of preview. Expand in Data Studio

ScanNet v2 meshes and segmentations

Selected original files from /data/scannet-v2, preserving their source filenames and directory layout. Files are uploaded without modification.

scans/: 1,513 scenes with all five files listed below.
scans_test/: 100 test scenes with mesh and metadata. The local source contains no label PLY, segment JSON, or aggregation JSON for these scenes.
scannetv2-labels.combined.tsv: the original label mapping.

Per-scene layout

scans/scene0123_01/
  scene0123_01_vh_clean_2.ply
  scene0123_01_vh_clean_2.labels.ply
  scene0123_01_vh_clean_2.0.010000.segs.json
  scene0123_01.aggregation.json
  scene0123_01.txt

Inventory

The selected scene files total 7,765 files and 20,880,177,310 bytes. files_manifest.jsonl records the relative path, byte count, and SHA-256 of every source file, including the label mapping. scene_manifest.json records each scene and its available and missing requested files.

Source and terms

Original dataset: ScanNet. See the official format documentation and ScanNet Terms of Use. This copy retains the original dataset terms.

Downloads last month: -