Ω-GAN: Object Manifold Embedding GAN for Image Generation by Disentangling Parameters into Pose and Shape Manifolds.
Yasutomo KawanishiDaisuke DeguchiIchiro IdeHiroshi MurasePublished in: ICPR (2020)
Keyphrases
- manifold embedding
- head pose estimation
- image generation
- pose estimation
- manifold learning
- d objects
- dimension reduction
- shape variations
- visual tracking
- parameter space
- shape descriptors
- partial occlusion
- random forests
- high resolution
- shape model
- low dimensional
- facial features
- feature extraction
- object tracking
- human motion
- viewpoint
- low dimensional manifolds
- object recognition
- shape features
- depth images
- high dimensional