Unsupervised Learning of Depth and Ego-Motion with Spatial-Temporal Geometric Constraints.
Anjie WangYongbin GaoZhijun FangXiaoyan JiangShanshe WangSiwei MaJenq-Neng HwangPublished in: ICME (2019)
Keyphrases
- spatial temporal
- geometric constraints
- ego motion
- camera motion
- epipolar geometry
- field of view
- depth information
- optical flow
- d scene
- stereo camera
- spatio temporal
- action recognition
- image sequences
- camera calibration
- motion parameters
- multi camera
- video shots
- scene structure
- dynamic scenes
- temporal information
- multiple images
- flow field
- moving camera
- optic flow
- motion field
- depth map
- moving objects
- video sequences
- human computer interaction
- image pairs
- computer vision
- facial expressions
- motion estimation
- velocity field
- image segmentation