drprojects/superpoint_transformer
Official PyTorch implementation of Superpoint Transformer introduced in [ICCV'23] "Efficient 3D Semantic Segmentation with Superpoint Transformer" and SuperCluster introduced in [3DV'24 Oral] "Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering"
PythonMIT
Issues
- 2
How to train in different devices?
#123 opened - 1
- 2
Training the model on my own dataset
#121 opened - 2
how to get the superpoints of each point
#120 opened - 3
shapenet dataset
#119 opened - 2
- 3
Running from a single inference script
#117 opened - 3
Questions about parametrization
#116 opened - 10
Empty list on Scatter_nearest_neighbor
#115 opened - 3
- 3
- 6
Is my format of the Scannetv2 wrong?
#112 opened - 9
Error when installing pgeof.
#111 opened - 1
Can the model run on smaller GPU (6GB) ?
#110 opened - 25
- 1
Norm_index
#108 opened - 10
- 3
Training custom dataset with all points
#106 opened - 1
Transformer_blocks architecture in SPT
#105 opened - 1
- 1
xy_tiling setting when process own data
#103 opened - 18
- 4
- 1
Is it possible to have more input channels?
#100 opened - 8
FRNN - RuntimeError: Unknown layout
#99 opened - 5
Issue in running demo
#98 opened - 11
- 2
- 2
No module named 'frnn'
#94 opened - 1
- 0
- 9
- 1
- 1
- 2
- 2
panoptic results
#87 opened - 3
About the decoder process
#86 opened - 2
there is no branch
#85 opened - 3
there are many strange problems
#84 opened - 4
- 1
about transformer block
#82 opened - 1
Superpoint generation
#81 opened - 3
S3DIS Fold 5 and 6 fold differences
#80 opened - 1
Input irregularly shaped point clouds
#79 opened - 32
Full resolution position
#78 opened - 17
Fail to install FRNN
#77 opened - 8
process got killed when generating superpoint, probably memory leak by parallel_cut_pursuit package
#76 opened - 4
- 2
file missing in v2 commit?
#74 opened - 3
the number of fold
#73 opened