Unsupervised Learning of Geometry From Videos With Edge-Aware Depth-Normal Consistency.
Zhenheng YangPeng WangWei XuLiang ZhaoRamakant NevatiaPublished in: AAAI (2018)
Keyphrases
- unsupervised learning
- supervised learning
- depth map
- depth information
- edge detection
- edge detector
- dimensionality reduction
- video clips
- video frames
- model selection
- edge information
- three dimensional
- semi supervised
- scene geometry
- stereoscopic video
- feature selection
- weighted graph
- video analysis
- video data
- training data
- key frames
- stereo vision
- video sequences
- reinforcement learning
- data sets
- deep architectures
- camera lens