I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image.
Gyeongsik MoonKyoung Mu LeePublished in: ECCV (7) (2020)
Keyphrases
- input image
- image content
- image data
- image features
- image classification
- template matching
- image analysis
- single image
- multiscale
- low level
- image representation
- estimation algorithm
- test images
- vector field
- segmentation method
- image segmentation
- image retrieval
- image pixels
- motion estimation
- high resolution
- image matching
- multiresolution
- similarity measure
- region of interest
- image processing