VDIGPKU/CBNetV2

Issues

• EVA-02

  #64 opened 2 months ago by JarvisKevin
  0

• "BatchNorm eval trick" in CBNet question

  #62 opened a year ago by zlenyk
  0

• Batchsize and lr of HTC+Swin-L

  #50 opened 3 years ago by BinZhu-ece
  2

• Slow inference speed with default config

  #23 opened 2 years ago by imyhxy
  1

• Welcome update to OpenMMLab 2.0

  #61 opened 2 years ago by vansin
  0

• backbone code for CB-HRNetv2pw32 and config files

  #60 opened 2 years ago by HaruHonda
  0

• Training custom dataset out of memory

  #58 opened 2 years ago by ttrungtin2910
  1

• i got this error when training cbnetv2, but It's normal when I train other models

  #57 opened 3 years ago by wzr0108
  11

• CBNetV2 installation on colab

  #59 opened 2 years ago by mohamedettebayo
  0

• FileNotFoundError: [Errno 2] No such file or directory: 'data/coco/stuffthingmaps/train2017/xyz.png'

  #27 opened 3 years ago by Resham-Sundar
  3

• How to disable semantic mask since I only want to do object detection?

  #42 opened 3 years ago by Icecream-blue-sky
  1

• 请问下注意力图是怎么实现的？

  #56 opened 3 years ago by XuKer
  0

• Could you please make sure running successfully before open source?

  #55 opened 3 years ago by liming-ai
  2

• so many bugs, why public...I can't even train existing model

  #43 opened 3 years ago by Icecream-blue-sky
  4

• About DB-Swin-B model, I tried to reimplementate it and get 55.2 mAP

  #45 opened 3 years ago by JR-Wang
  13

• How to use three backbone?

  #30 opened 3 years ago by Klawens
  2

• About inference speed

  #24 opened 3 years ago by WongKinYiu
  2

• Inference phase w.out labels

  #25 opened 3 years ago by aliceinland
  2

• 请问，已训练好的模型，想使用cbnet 应该怎么做 ？ 比如说 我想使用两个 backbone 三个backbone

  #32 opened 3 years ago by lgf821793883
  3

• Single GPU training

  #37 opened 3 years ago by Ali-Abolfathi
  1

• inference speed is too slow : 360*640 Picture takes 1s

  #20 opened 3 years ago by ahong007007
  2

• Does it work on pytorch1.9.0?

  #54 opened 3 years ago by rainbow-xiao
  0

• ERROR 404 NOT FOUND - get_started.md

  #53 opened 3 years ago by Quentinmht
  1

• ZeroDivisionError: float division by zero

  #52 opened 3 years ago by xingkongliang
  0

• Possible bug in mmdet modification?

  #51 opened 3 years ago by maitreygram
  0

• Reproduce CBNetV2 with other models

  #49 opened 3 years ago
  0

• If I understood it right all the algorithms are mask based right except for Faster-Rcnn model, right? Can I use the cascade-mask-rcnn model for just object detection, i.e., to train the model without segmentation annotations?

  #46 opened 3 years ago by CharlesNJ
  1

• Where is the source code of `CBSwinTransformerv2`?

  #48 opened 3 years ago by mangoyuan
  1

• Reproduce DB-Swin-Large models

  #47 opened 3 years ago by cailk
  0

• model loading issues

  #36 opened 3 years ago by explorer359
  4

• What is actual pytorch, cuda , mmdet,mmcv version used?

  #44 opened 3 years ago by VikasRajashekar
  1

• Loss goes nan when training dual swin-base

  #39 opened 3 years ago by seanzhuh
  2

• RecursionError: maximum recursion depth exceeded while calling a Python object

  #29 opened 3 years ago by CharlesNJ
  8

• Is there a way to only detect the classes that I need with the pre-trained model, example car and truck?

  #35 opened 3 years ago by CharlesNJ
  2

• Does this model take different sizes of images as i/p? Or does all the images have to be of the same WxH?

  #34 opened 3 years ago by CharlesNJ
  2

• CBSwinTransformer is not in the models registry

  #18 opened 3 years ago by ahong007007
  5

• CBNet Paper Layers composition

  #21 opened 3 years ago by aliceinland
  1

• KeyError: "CascadeRCNN: 'CBRes2Net is not in the models registry'"

  #17 opened 3 years ago by fzfs
  1

• Questions about fused_semantic_head

  #41 opened 3 years ago by JR-Wang
  1

• About train on custom dataset

  #40 opened 3 years ago by linyangsdu
  0

• 为什么我测试 所有ap都为0

  #33 opened 3 years ago by lgf821793883
  1

• how to train DB-Swin-L?

  #38 opened 3 years ago by fuweifu-vtoo
  0

• KeyError: 'CBSwinTransformer is not in the models registry'

  #31 opened 3 years ago by CharlesNJ
  10

• How to show the results images?

  #26 opened 3 years ago by RainbowSun11Q2H
  2

• How to perform instance segmentation using CBNetV2?

  #28 opened 3 years ago by harrytrinh96
  0

• How to train on a custom dataset

  #15 opened 3 years ago by aliceinland
  10

• can you release the DB-Swin-L of training-config file ?

  #22 opened 3 years ago by larsoncs
  0

• got it

  #14 opened 3 years ago by larsoncs
  0

• FileNotFoundError: [Errno 2] No such file or directory: 'data/coco/train2017/000000256364.png'

  #19 opened 3 years ago by LLsmile
  0

• Can pre-trained COCO models be made available?

  #16 opened 3 years ago by morgankohler
  0