Panoptic Nuscenes: A Large-Scale Benchmark for LiDAR Panoptic Segmentation and Tracking.
Whye Kit FongRohit MohanJuana Valeria HurtadoLubing ZhouHolger CaesarOscar BeijbomAbhinav ValadaPublished in: IEEE Robotics Autom. Lett. (2022)
Keyphrases
- motion segmentation
- real world
- image segmentation
- berkeley segmentation dataset
- segmentation algorithm
- deformable templates
- level set
- automatic tracking
- multiple objects
- particle filter
- segmentation method
- motion cues
- foreground background separation
- automatic initialization
- quantitative evaluation
- kalman filter
- multiscale
- object segmentation
- occlusion reasoning
- high resolution
- point cloud
- real time
- image sequences
- accurate localization
- object tracking
- particle filtering
- shape prior
- region growing
- visual tracking
- bounding box
- motion tracking
- fully unsupervised
- multiple cues
- active contours
- graph cuts
- fully automatic
- automatic segmentation