Voxel-MAE: Masked Autoencoders for Pre-training Large-scale Point Clouds.
Chen MinDawei ZhaoLiang XiaoYiming NieBin DaiPublished in: CoRR (2022)
Keyphrases
- point cloud
- surface reconstruction
- structure from motion
- point cloud data
- feedforward neural networks
- triangular mesh
- laser scanner
- data sets
- training set
- lidar data
- iterative closest point
- restricted boltzmann machine
- denoising
- stereo camera
- architectural scenes
- normal vectors
- point sets
- multi view
- nearest neighbor
- d objects
- query processing
- support vector
- three dimensional