Intersection Detector - Training Dataset


Abstract: 
In this work, we propose a probabilistic method for detecting and classifying urban road intersections from a moving vehicle. In order to evaluate our approach we demonstrate its capabilities in intersection topology classification with respect to the 7 intersection patterns shown in the figure below. We manually annotated the following KITTI sequences to create the intersection topology ground truth, associating each frame of each sequence to the corresponding topology. We also introduced a flag (referred to as the "crossing" flag), which indicates whether the vehicle is inside an intersection.

type_0 type_1 type_2 type_3 type_4 type_5 type_6
0 1 2 3 4 5 6

Road surface dataset
The road surface training dataset is based on the work of the Torr Vision Group, which can be found at the following link: Urban 3D Semantic Modelling Using Stereo Vision.

We labelled again their dataset using the following 3 classes only:

  • road
  • sidewalk / pavement
  • other, i.e., everything else

Moreover, we additionally labelled 10 images from the original KITTI sequences. The new images contain only intersection areas. Thus, the resulting dataset consists of the original 70 images (25 from the training set + 45 from their test set) plus 10 images.

The resulting dataset can be downloaded from the following  iralab-intersection-training.zip (37 downloads)

We tried also to match the original TORR framenames with the corresponding KITTI dataset. The matching is included in the aforementioned link and in the following table:

Dataset + frame name name in torr + ours
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002970.png 08_001870.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/08_000896.png 08_000896.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000003381.png 0000003381.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000000187.png 0000000187.png
2011_09_30/2011_09_30_drive_0027_sync/image_02/data/0000000510.png 07_000510.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001317.png 08_000217.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001262.png 08_000162.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001100.png 08_000000.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000004121.png 0000004121.png
2011_09_30/2011_09_30_drive_0027_sync/image_02/data/0000000190.png 07_000190.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002953.png 08_001853.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002606.png 08_001506.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000004412.png 08_003312.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000002595.png 0000002595.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003648.png 08_002548.png
2011_09_30/2011_09_30_drive_0018_sync/image_02/data/0000000095.png 05_000095.png
2011_09_30/2011_09_30_drive_0020_sync/image_02/data/0000000845.png 06_000845.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001540.png 08_000440.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003333.png 08_002233.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002998.png 08_001898.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000004407.png 0000004407.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000004166.png 08_003066.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001220.png 08_000120.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002198.png 08_001098.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000004745.png 08_003645.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001738.png 08_000638.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002259.png 08_001159.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001703.png 08_000603.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003470.png 08_002370.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000003937.png 0000003937.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002897.png 08_001797.png
2011_09_30/2011_09_30_drive_0027_sync/image_02/data/0000000438.png 07_000438.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000003514.png 0000003514.png
2011_09_30/2011_09_30_drive_0020_sync/image_02/data/0000000664.png 06_000664.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002078.png 08_000978.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003063.png 08_001963.png
2011_09_30/2011_09_30_drive_0027_sync/image_02/data/0000000169.png 07_000169.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001678.png 08_000578.png
2011_09_30/2011_09_30_drive_0027_sync/image_02/data/0000000090.png 07_000090.png
2011_09_30/2011_09_30_drive_0018_sync/image_02/data/0000000853.png 05_000853.png
2011_09_30/2011_09_30_drive_0020_sync/image_02/data/0000000472.png 06_000472.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003456.png 08_002356.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002572.png 08_001472.png
2011_09_30/2011_09_30_drive_0018_sync/image_02/data/0000000754.png 05_000754.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002481.png 08_001381.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001615.png 08_000515.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002542.png 08_001442.png
2011_09_30/2011_09_30_drive_0020_sync/image_02/data/0000000945.png 06_000945.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000000106.png 0000000106.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003364.png 08_002264.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000001186.png 08_000086.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000002422.png 08_001322.png
2011_09_30/2011_09_30_drive_0028_sync/image_02/data/0000003115.png 08_002015.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000002854.png 0000002854.png
2011_10_03/2011_10_03_drive_0027_sync/image_02/data/0000000368.png 0000000368.png

The following are the 10 new images we inserted in our training set.

0000000106.png
0000000187.png
0000000368.png
0000002595.png
0000002854.png
0000003381.png
0000003514.png
0000003937.png
0000004121.png
0000004407.png

0000004407 0000004407
0000004121 0000004121
0000003937 0000003937
0000003514 0000003514
0000003381 0000003381
0000002854 0000002854
0000002595 0000002595
0000000368 0000000368
0000000187 0000000187
0000000106 0000000106