CATS 2: Color And Thermal Stereo Scenes with Semantic Labels

The CATS dataset introduced a new set of diverse indoor and outdoor scenes with ground truth disparity information for testing stereo matching algorithms in color and thermal imagery. These scenes included nighttime, foggy, low light, and complex lighting in scenes. To extend the usefulness of the CATS dataset we add pixeland instance-level semantic labels. This includes labels for both color and thermal imagery, and the labels also apply to 3D point clouds as a result of the existing 2D-3D alignment. We compare the new CATS 2.0 dataset against other similar datasets and show it is similar in scope to the KITTI-360 and WildDash datasets, but with the addition of both thermal and 3D information. Additionally, we run a benchmark pedestrian detection algorithm on a set of scenes containing pedestrians.

[1]  Scott Sorensen,et al.  CATS: A Color and Thermal Stereo Benchmark , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Marc Pollefeys,et al.  Semantic3D.net: A new Large-scale Point Cloud Classification Benchmark , 2017, ArXiv.

[3]  Oliver Zendel,et al.  WildDash - Creating Hazard-Aware Benchmarks , 2018, ECCV.

[4]  Sebastian Ramos,et al.  The Cityscapes Dataset for Semantic Urban Scene Understanding , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Franz Rottensteiner,et al.  ISPRS Test Project on Urban Classification and 3D Building Reconstruction: Evaluation of Building Reconstruction Results , 2009 .

[6]  Andreas Geiger,et al.  Vision meets robotics: The KITTI dataset , 2013, Int. J. Robotics Res..

[7]  Pietro Perona,et al.  Microsoft COCO: Common Objects in Context , 2014, ECCV.

[8]  Andreas Geiger,et al.  Are we ready for autonomous driving? The KITTI vision benchmark suite , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[9]  Chengyang Li,et al.  Multispectral Pedestrian Detection via Simultaneous Detection and Segmentation , 2018, BMVC.

[10]  Namil Kim,et al.  Multispectral pedestrian detection: Benchmark dataset and baseline , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Jannik Fritsch,et al.  A new performance measure and evaluation benchmark for road detection algorithms , 2013, 16th International IEEE Conference on Intelligent Transportation Systems (ITSC 2013).

[12]  Jun Zhang,et al.  Benchmark datasets for 3D computer vision , 2014, 2014 9th IEEE Conference on Industrial Electronics and Applications.

[13]  Andreas Geiger,et al.  Augmented Reality Meets Computer Vision: Efficient Data Generation for Urban Driving Scenes , 2017, International Journal of Computer Vision.

[14]  Sebastian Ramos,et al.  The Cityscapes Dataset , 2015, CVPR 2015.