Multi-task human analysis in still images: 2D/3D pose, depth map, and multi-part segmentation.
Daniel SánchezMarc OliuMeysam MadadiXavier BaróSergio EscaleraPublished in: CoRR (2019)
Keyphrases
- depth map
- input image
- multi task
- segmentation method
- three dimensional
- image data
- image registration
- image classification
- pose estimation
- medical images
- multi task learning
- depth estimation
- image features
- image segmentation
- depth information
- level set
- multi view
- camera pose
- dense depth map
- data points
- active learning
- object recognition
- high quality