Megvii-BaseDetection/BEVDepth

Issues

• can't reproduce [bev_depth_lss_r50_256x704_128x128_24e_2key_ema]

  #168 opened a year ago by kimjh7669
  1

• Input type different & Memory usage during training

  #165 opened a year ago by hutchinsonian
  2

• map仅仅只有0.0003

  #201 opened 3 months ago by Hichengdong
  1

• BCE Loss or L1 Loss for Depth Loss

  #200 opened 4 months ago by DFLyan
  0

• How to deal with camera pixels without lidar point?

  #175 opened a year ago by klong121
  2

• Depth aggregation autocast

  #163 opened a year ago by gaguai
  2

• ValueError: not enough values to unpack (expected 10, got 1)

  #199 opened 5 months ago by YuhanZhou521
  0

• Data Set Processing

  #198 opened 5 months ago by YuhanZhou521
  0

• 请问bevdepth文章对应的是bev_depth_lss还是bev_depth_fusion_lss？

  #197 opened 5 months ago by kingofstu
  0

• voxel_pooling_train_ext.voxel_pooling_train_forward_wrapper

  #196 opened 5 months ago by kingofstu
  0

• The inference results of the MatrixVT model do not match the ground truth.

  #195 opened 6 months ago by wuhandashuaibi
  0

• How to use VoVnet-99 in BEVDepth

  #194 opened 7 months ago by altkddhfcjs
  0

• Where is Depth Refine Module?

  #193 opened 7 months ago by cxl-ustb
  0

• How can I get the results you reported in the programme?

  #192 opened 8 months ago by hxyHXY612
  0

• 监督的bce-loss

  #191 opened 8 months ago by downpeopleup
  0

• 版本修改问题

  #190 opened 8 months ago by iphanazi
  0

• train_code_error

  #189 opened 8 months ago by wangzhenhao123
  0

• big gap in mAP

  #188 opened 9 months ago by csh-utopia
  0

• reproduce the code using mmdet/3d

  #187 opened 9 months ago by Li-Whasaka
  0

• so ugly code

  #186 opened 10 months ago by hopef
  0

• The Detetion Result

  #185 opened 10 months ago by goahence
  0

• lr scheduler `MultiStepLR` doesn't follow PyTorch's LRScheduler API.

  #184 opened 10 months ago by ChirsHuanhuan
  0

• replace backbone with Vit

  #183 opened 10 months ago by Alexanderisgod
  0

• Question about MatrixVT

  #182 opened a year ago by ZhouZijie77
  0

• BEV size must be square

  #181 opened a year ago by GYJamor
  0

• depth aggregation module in trian not in infer

  #179 opened a year ago by wutheringcoo
  0

• How to visualize a bev map?

  #180 opened a year ago by sidiangongyuan
  0

• mmcv mmdet mmseg 版本问题

  #171 opened a year ago by steven9046
  3

• How to test after training??

  #158 opened 2 years ago by soohyoen
  1

• sensor2sensor_mats doesn't seem to be used.

  #178 opened a year ago by Camellia-hz
  1

• Why Depthnet need take bda into account?

  #177 opened a year ago by pdoublerainbow
  0

• gen_info error

  #176 opened a year ago by Alexanderisgod
  0

• error: no instance of overloaded function "atomicAdd" matches the argument list

  #159 opened a year ago by yukaizhou
  5

• Memory Peak

  #174 opened a year ago by thom966
  0

• what the accuracy of your depth map fusion model?

  #173 opened a year ago by BoomSky0416
  0

• python setup.py develop Error

  #157 opened 2 years ago by yukaizhou
  7

• Performance of MatrixVT

  #166 opened a year ago by zhenghw2000
  3

• error detected in the compilation of "/tmp/tmpxft_00005518_00000000-6_voxel_pooling_train_forward_cuda.cpp1.ii

  #170 opened a year ago by steven9046
  2

• 推理速度

  #172 opened a year ago by aqr1961
  0

• Deployment of BEVDepth on Nvidia board with onnx conversion

  #169 opened a year ago by zhoumu1990
  0

• Unfreeze backbone network

  #160 opened a year ago by thom966
  1

• inference with depth gt

  #167 opened a year ago by kimjh7669
  1

• BEVDepth perform very bad when training with batch size 1!

  #164 opened a year ago by cskkxjk
  2

• BEV grid resolution

  #154 opened 2 years ago by yukaizhou
  1

• Welcome update to OpenMMLab 2.0

  #162 opened a year ago by vansin
  0

• Questions for BEV features and voxel pooling

  #161 opened a year ago by DianCh
  0

• Some questions about code matrixvt.py

  #152 opened 2 years ago by wuwangzhuanwan
  6

• key_idxes = [-1] in BEVDepth?

  #156 opened 2 years ago by lukas-staecker
  0

• question about the setting of EMA

  #155 opened 2 years ago by zen-d
  0

• pkl file used in visuasualization code does not exist

  #153 opened 2 years ago by djuneja17
  0