Unsupervised Disentanglement of Pose, Appearance and Background from Images and Videos.
Aysegul DundarKevin J. ShihAnimesh GargRobert PottorfAndrew TaoBryan CatanzaroPublished in: CoRR (2020)
Keyphrases
- input image
- image features
- bounding box
- visual appearance
- static images
- image classification
- visual patterns
- pixel wise
- probe image
- camera viewpoint
- viewing angle
- image retrieval
- three dimensional
- object recognition
- keypoints
- d objects
- video data
- video sequences
- cluttered background
- background clutter
- complex background
- illumination conditions
- partial occlusion
- lighting conditions
- dynamic textures
- appearance model
- pose estimation
- shape variations
- target object
- object shape
- object appearance
- calibrated cameras
- pose space
- test images