SamsungLabs/fcaf3d

Issues

• problem when install mmdet3d

  #53 opened a month ago by Pixie8888
  2

• How to solve the error reported during the training trial

  #52 opened a year ago by nss6planner
  3

• Welcome update to OpenMMLab 2.0

  #50 opened a year ago by vansin
  1

• Training with Custom Dataset

  #49 opened 2 years ago by Shromm-Gaind
  1

• FCAF3D evaluation result mAP question

  #48 opened 2 years ago by wenchened
  2

• run pcd_demo.py

  #36 opened 2 years ago by PengboLi1998
  3

• Generalization of FCAF3D

  #45 opened 2 years ago by Eaphan
  2

• terminate called after throwing an instance of 'c10::CUDAError'

  #44 opened 2 years ago by zyrant
  1

• Concatenate different feature levels to one level before passing data to the heads

  #43 opened 2 years ago by FabianSchuetze
  2

• Can MinkowskiGenerativeConvolutionTranspose from fcaf3d_neck_with_head.py replaced?

  #33 opened 2 years ago by YJYJLee
  4

• Performance on SUN RGB-D

  #42 opened 2 years ago by Sharpiless
  5

• resume-from

  #41 opened 2 years ago by rgrbic
  0

• visualization question

  #31 opened 2 years ago by PengboLi1998
  7

• The error about training the S3DIS dataset

  #40 opened 2 years ago by week0425
  5

• Performance on ScanNet

  #39 opened 2 years ago by WWW2323
  4

• log information

  #35 opened 2 years ago by PengboLi1998
  1

• output all predict bbox coordinate position

  #34 opened 2 years ago by PengboLi1998
  4

• mAP result question

  #37 opened 2 years ago by chence17
  2

• another version fcaf3d

  #28 opened 2 years ago by Heroism502
  3

• train.py: error: unrecognized arguments: --load-from fully_conv_s3dis.pth

  #32 opened 2 years ago by PengboLi1998
  1

• RuntimeError: NCCL error in: /opt/conda/conda-bld/pytorch_1603729009598/work/torch/lib/c10d/ProcessGroupNCCL.cpp:784, invalid usage, NCCL version 2.7.8

  #29 opened 2 years ago by PengboLi1998
  1

• ModuleNotFoundError: No module named 'mmdet3d.ops.pcdet_nms'

  #30 opened 2 years ago by PengboLi1998
  0

• How can I generate per point classification?

  #27 opened 2 years ago by sylyt62
  4

• Training on KITTI

  #12 opened 3 years ago by abhigoku10
  4

• How does demo predict gravity direction without calibration data?

  #26 opened 3 years ago by joshiaLee
  2

• mAP depending on score_thr

  #25 opened 3 years ago by joshiaLee
  2

• Some question on code (About the origin in class DepthInstance3DBoxes )

  #24 opened 3 years ago by wtt0213
  5

• how can I increase small objects'mAP?

  #23 opened 3 years ago by joshiaLee
  5

• I encounter the error KeyError: 'gt_bboxes_3d' when i try to use visualize_results.py

  #21 opened 3 years ago by snow-flakes1215
  2

• I follow the step of file 'read me' except that i use only a single gpu and it errors: "RuntimeError: merge_sort: failed to synchronize: cudaErrorIllegalAddress: an illegal memory access was encountered"

  #19 opened 3 years ago by BrotherHappy
  7

• I'd like to detect small objects

  #20 opened 3 years ago by joshiaLee
  3

• I run the trained model on a single data and encountered buffer size issue

  #22 opened 3 years ago by joshiaLee
  3

• how should I understand results?

  #18 opened 3 years ago by joshiaLee
  2

• Nothing happened when I visualized my latest.pth

  #17 opened 3 years ago by joshiaLee
  6

• some issues for the evaluation results of fcaf3d

  #15 opened 3 years ago by zschanghai
  4

• Can you please share the docker file for CUDA 11.2

  #16 opened 3 years ago by giangdip2410
  2

• Alternative Voxel Based Backbones?

  #14 opened 3 years ago by stanleyshly
  3

• Merge with main mmdetection3d repo?

  #13 opened 3 years ago by stanleyshly
  2