VG-VAE: A Venatus Geometry Point-Cloud Variational Auto-Encoder.
Tejas AnvekarRamesh Ashok TabibDikshit HegdeUma MudengudiPublished in: CVPR Workshops (2022)
Keyphrases
- point cloud
- multi view stereo
- cad model
- surface reconstruction
- structure from motion
- image segmentation
- stereo camera
- point sets
- triangular mesh
- laser scanner
- dense reconstruction
- point cloud data
- iterative closest point
- photorealistic
- three dimensional
- optical flow
- optical flow computation
- bit rate
- real world objects
- computer vision
- motion estimation
- visual hull
- geometric constraints
- surface representation
- nearest neighbor