Michelangelo: Conditional 3D Shape Generation based on Shape-Image-Text Aligned Latent Representation.
Zibo ZhaoWen LiuXin ChenXianfang ZengRui WangPei ChengBin FuTao ChenGang YuShenghua GaoPublished in: NeurIPS (2023)
Keyphrases
- object shape
- image representation
- single image
- geometric shapes
- shape estimation
- image features
- multiscale
- input image
- image classification
- object shapes
- primal sketch
- spatial layout
- image structure
- spatial arrangement
- image measurements
- image segmentation
- surface shape
- spatial relations
- relative position
- segmentation method
- chain code
- vector field
- image content
- deformable template
- hand drawn
- energy functional
- segmentation algorithm
- scale space
- spatial configuration
- shape representation
- invariant descriptors
- text graphics
- shape retrieval
- geometric information
- reconstruction method
- fourier descriptors
- invariant features
- object boundaries
- energy function
- multi view
- edge detection
- d objects
- high resolution
- image retrieval
- three dimensional