BKinD-3D: Self-Supervised 3D Keypoint Discovery from Multi-View Videos.
Jennifer J. SunLili KarashchukAmil DravidSerim RyouSonia FereidooniJohn C. TuthillAggelos K. KatsaggelosBingni W. BruntonGeorgia GkioxariAnn KennedyYisong YuePietro PeronaPublished in: CVPR (2023)
Keyphrases
- multi view
- dynamic scenes
- multi view video sequences
- multiple views
- single view
- view invariant
- three dimensional
- camera calibration
- d objects
- multiple cameras
- keypoints
- depth map
- video sequences
- video content
- free viewpoint
- semi supervised
- view synthesis
- video frames
- multiple viewpoints
- co training
- multi view stereo
- surface reconstruction
- multi view images
- image features
- stereoscopic video
- multi view clustering
- visual hull
- depth video
- machine learning
- depth information
- video surveillance
- multi view reconstruction