A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting Hand Pose Estimation from a Single RGB Image.
Changlong JiangYang XiaoCunlin WuMingyang ZhangJinghong ZhengZhiguo CaoJoey Tianyi ZhouPublished in: CoRR (2023)
Keyphrases
- image data
- single image
- input image
- image analysis
- image representation
- distribution network
- image features
- fuzzy logic
- image segmentation
- image retrieval
- edge detection
- image content
- multiscale
- image classification
- segmentation method
- test images
- colour space
- image regions
- power transformers
- region of interest
- rgb color
- pixel values
- color histogram
- color channels
- spatial coordinates
- neural network
- image collections
- fault diagnosis
- color space
- gray level
- image quality
- peer to peer
- low level
- computer vision