End-to-End Vision-to-Motion Model with Auxiliary Segmentation Module for Indoor Navigation.
Weili ChenChao JiaShuai HePublished in: ICDLT (2019)
Keyphrases
- end to end
- motion model
- motion segmentation
- indoor navigation
- multibody
- optical flow estimation
- motion estimation
- optical flow
- particle filter
- image sequences
- data association
- motion field
- computer vision
- motion parameters
- multiple motion models
- image segmentation
- image motion
- level set
- vision system
- flow field
- video sequences
- multiple objects
- multiscale
- high quality
- real time
- shape prior
- moving objects
- congestion control
- depth ordering
- structured environments
- machine learning