DiffBody: Diffusion-based Pose and Shape Editing of Human Images.
Yuta OkuyamaYuki EndoYoshihiro KanamoriPublished in: CoRR (2024)
Keyphrases
- image data
- input image
- human subjects
- texture information
- image features
- object shape
- multiple images
- diffusion coefficient
- shape variations
- tensor field
- improve recognition accuracy
- image retrieval
- object recognition
- lighting conditions
- appearance variations
- three dimensional
- human head
- articulated pose
- lighting direction
- position and orientation
- partial occlusion
- shape model
- image matching
- edge detection
- illumination variations
- aspect ratio
- surface shape
- image editing
- prostate segmentation
- keypoints
- scalar valued
- image registration
- image segmentation