Download

Open Resource

  • POSS-TERRAIN: A Vision-Proprioception terrain Dataset under various lighting conditions. (Page Link )
  • TCPOSS: A terrain classification comprising terrain data of 10 categories in Peking University, reflecting rich intra-class diversity. (Page Link )
  • Datasets for the 3D Semantic Segmentation to the Open World (3DOW) Challenge: SubKITTI and AugKITTI (Page Link )
  • POSS Odometry: A Visual and LiDAR Odometry Dataset in Campus Environment (Page Link )
  • SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic Instances (Page Link )
  • Dataset for semantic segmentation of 3d lidar data in dynamic scene using semi-supervised learning(2019 Mei T-ITS)(Zip 176MB, without raw LiDAR point cloud; Zip 2.16GB, with raw LiDAR point cloud)
The POSS is part of the School of EECS, Peking University..