ShapeClipper: Scalable 3D Shape Learning from Single-View Images via Geometric and CLIP-Based Consistency.
Zixuan HuangVarun JampaniAnh ThaiYuanzhen LiStefan StojanovJames M. RehgPublished in: CVPR (2023)
Keyphrases
- single view
- multi view
- three dimensional
- polyhedral objects
- calibrated cameras
- learning process
- input image
- learning algorithm
- man made objects
- perspective projection
- aerial images
- image data
- vanishing points
- single image
- geometric information
- co training
- geometric structure
- active learning
- viewpoint
- object recognition
- training data
- computer vision