xinge008/Cylinder3D
Rank 1st in the leaderboard of SemanticKITTI semantic segmentation (both single-scan and multi-scan) (Nov. 2020) (CVPR2021 Oral)
PythonApache-2.0
Issues
- 1
Point wise refinement (backreference)
#125 opened - 1
- 8
- 2
How to visualize the results of the demo?
#122 opened - 5
spconv error shape in SubMConv3d
#121 opened - 0
- 3
Any idea to make the inference faster ?
#119 opened - 2
- 0
- 6
- 2
point-wise refine moduls
#114 opened - 3
a multi-scan pretrained model?
#112 opened - 2
pretraind model
#111 opened - 2
如何使用自己标注的数据集进行训练?
#110 opened - 2
CUDA out of memory??
#109 opened - 1
Rosbag dataset
#108 opened - 23
- 1
Lane
#106 opened - 2
'std::bad_alloc' error while trying the demo
#105 opened - 1
Docker build failing at step 30
#103 opened - 1
- 0
训练集加载后长度为零
#101 opened - 1
value of "label_size? thanks!
#100 opened - 1
- 0
iou calculation about the
#98 opened - 9
spconv 2.1.6 error shape in SubMConv3d
#97 opened - 1
- 1
Question about details
#95 opened - 1
about test.py?
#94 opened - 11
- 1
understanding about AU part?
#92 opened - 1
- 7
- 5
使用GPU训练的问题以及如何使用pretrained model测试
#89 opened - 9
Difference in SemanticKITTI val mIoU
#88 opened - 1
Missing license
#87 opened - 2
nuscenes dataset
#86 opened - 0
nuscenes dataset
#85 opened - 0
3D polar convolution
#84 opened - 2
Wrong prediction of released nuscences Model
#83 opened - 3
location of checkpoint file
#82 opened - 0
location of checkpoint file
#81 opened - 2
- 3
- 2
train loss
#78 opened - 5
RuntimeError: cudaErrorIllegalAddress
#77 opened - 2
- 1
Panotic Segmentation code release
#75 opened - 2
NuScenes multiple sweeps
#74 opened - 16
Visualization
#73 opened