Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image.
Denis TomèChris RussellLourdes AgapitoPublished in: CoRR (2017)
Keyphrases
- single image
- deep learning
- multiple images
- deep belief networks
- wavelet transform
- image pairs
- shape recovery
- d scene
- super resolution
- pose estimation
- shape from shading
- indoor scenes
- spherical harmonics
- light source
- scene geometry
- high quality
- partially occluded
- three dimensional
- restricted boltzmann machine
- scene understanding
- human pose
- multiscale
- shadow removal
- image processing