TencentYoutuResearch/CrowdCounting-P2PNet

Issues

• 关于UCF-QNRF数据集训练问题

  #91 opened a month ago by susu-source
  2

• 给各位省点时间，看到我这条issue就不要继续折腾这个repo了

  #81 opened 7 months ago by SalenGit
  4

• 关于QNRF等数据集的结果问题

  #69 opened a year ago by yangtle
  40

• How to Replace backbone

  #70 opened a year ago by DragonBoyL
  38

• 为什么我增加了训练数据集数量后，无法训练？

  #79 opened 8 months ago by wangwangteam
  10

• P2PNet qq交流群：790745540，互相帮助，解决bug问题

  #90 opened 2 months ago by nice98k
  0

• P2PNet qq交流群： 790745540

  #89 opened 2 months ago by nice98k
  0

• Change backbone

  #86 opened 5 months ago by paolopepe00
  1

• 跑qnrf和nwpu数据集，报显存，有人解决过这个问题吗？ 显卡3090 24G

  #88 opened 2 months ago by nice98k
  0

• 效果很差

  #87 opened 2 months ago by yxl502
  0

• 在SHHB, QNRF, NWPU数据集上的参数设置

  #51 opened 2 years ago by msn199959
  5

• How to convert model to onnx

  #85 opened 5 months ago by TruongNV-hut
  0

• 这个模型如何更换backbone骨干网络？

  #83 opened 6 months ago by nice98k
  0

• 训练不收敛

  #84 opened 6 months ago by lp-094
  5

• 定位性能汇报

  #76 opened 10 months ago by streamer-AP
  3

• Facing issues in collate_fn_crowd(batch) function of util.misc.py

  #82 opened 7 months ago by abinashlingank
  1

• 求助

  #73 opened a year ago by weeoou
  1

• Oral的paper代码写的这么差，头次见

  #63 opened a year ago by ShowMeTheCodePLZ
  5

• 维度128*128为什么可以输入VGG

  #78 opened 9 months ago by Minsky520
  1

• requirements.txt版本？

  #67 opened a year ago by zhushaokun123
  1

• "train.list" file

  #77 opened 9 months ago by aliva-ksu
  1

• 跑不通代码

  #75 opened 10 months ago by liangguomin
  0

• 如何添加指标R2和RE

  #74 opened 10 months ago by Scobbb
  0

• 如何将数据集中的mat文件转换为TXT文件

  #72 opened a year ago by jone222
  2

• shanghaitechA dataset

  #50 opened 2 years ago by umair-shah
  3

• Hardcoded pyramid_levels causing runtime error with stride customization

  #71 opened a year ago by hanoglu
  0

• Softmax Normalization Missing in ClassificationModel

  #66 opened a year ago by Younger330
  2

• 代码有误

  #61 opened 2 years ago by jhxiang
  1

• 提供的权重测试指标无法达到论文报告的水平（Deviation from expected performance on SHTechA dataset）

  #65 opened a year ago by ocean146
  1

• regression = self.regression(features_fpn[1]) * 100

  #68 opened a year ago by SherlockHolmes221
  1

• Accuracy of UCF-QNRF

  #44 opened 2 years ago by 1286710929
  3

• 训练过程中，显存忽高忽低

  #36 opened 3 years ago by yangtle
  4

• 为什么越大的越近的，检测不到？

  #39 opened 3 years ago by xiakj
  4

• problem on random_crop when training (train.py 運行到random_crop時出問題)

  #53 opened 2 years ago by SingCheng
  3

• RuntimeError when doing image resize in run_test.py. please help ~

  #46 opened a year ago by slent310
  2

• the input image size

  #64 opened a year ago by ycbilge
  0

• [Bug]: key error in weight_dict & losses

  #48 opened 2 years ago by zhiyuanyou
  1

• Trained weight on UCF-QNRF dataset

  #62 opened 2 years ago by RuiDream
  0

• Dont use full GPU usage in A100

  #60 opened 2 years ago by VietPT3502
  0

• Parameters for Shanghai Tech part B

  #59 opened 2 years ago by Sssanek
  0

• 您好，我也是一样的情况，在出现了Namespace(backbone='vgg16_bn', gpu_id=0, line=2, output_dir='./logs/', row=2, weight_path='./weights/SHTechA.pth')之后，后面就没有反应了。不知道有没有顺利解决这个问题的

  #49 opened 2 years ago by fatakWang
  2

• Problem on using vgg11_bn to run `run_test.py`

  #57 opened 2 years ago by chilldenaya
  1

• How to run model on video input?

  #52 opened 2 years ago by mariosconsta
  2

• Is there any specific reason as to why VGG-16 network is chosen for feature extraction?

  #47 opened 2 years ago by bit-scientist
  0

• Can the performance of the best_mae.pth reach 51.xx?

  #45 opened 2 years ago by msn199959
  1

• annotation data

  #43 opened 2 years ago by YN-e1023
  1

• distributed training

  #42 opened 2 years ago by shaunyuan22
  0

• Don't you plan to open source the computing method of nAP?

  #40 opened 2 years ago by zyweven
  0

• 怎么样才能导入视频流进行检测呢

  #38 opened 3 years ago by 1449362433
  0

• How to support video

  #37 opened 3 years ago by 1449362433
  0