MV-JAR: Masked Voxel Jigsaw and Reconstruction for LiDAR-Based Self-Supervised Pre-Training.
Runsen XuTai WangWenwei ZhangRunjian ChenJinkun CaoJiangmiao PangDahua LinPublished in: CoRR (2023)
Keyphrases
- three dimensional
- high resolution
- aerial imagery
- training samples
- lidar data
- cooperative learning
- training process
- training set
- supervised learning
- reconstruction method
- image reconstruction
- training phase
- data sets
- reconstruction process
- point cloud
- test set
- training examples
- collaborative learning
- semi supervised
- video sequences