zhongyy/Face-Transformer

Issues

• 无法安装依赖mxnet~=1.8.0.post0 与sklearn~=0.0 要求版本

  #40 opened 9 months ago by mgmpmd
  0

• What's the meaning of "The content of “property” file for “ms1m_retinaface” dataset is as follows: "93431,112,112""

  #39 opened a year ago by Cassieyy
  4

• Python version not mentioned

  #22 opened 3 years ago by aasharib
  3

• 你好 打算复现您的代码，请问这个怎么解决？ 报错（在 '__init__.py' 中找不到引用 'vit_face'）

  #24 opened 3 years ago by zydongmingyu
  3

• readme中预训练模型的权重是用哪些数据集训练的？

  #38 opened a year ago by bjzhb666
  0

• Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:1! (when checking argument for argument index in method wrapper_CUDA_scatter__value)

  #37 opened 2 years ago by azizyemen1
  0

• Is there no image normalization/pre-processing?

  #36 opened 2 years ago by arunsubk
  0

• Training days time

  #2 opened 4 years ago by khawar-islam
  5

• Source of DATA_ROOT?

  #12 opened 4 years ago by Martlgap
  2

• Training Log?

  #5 opened 4 years ago by XinWangg
  6

• Missing IJB-C dataset

  #14 opened 3 years ago by khawar-islam
  4

• Can i train this model with a dataset other than ms1m-retinaface

  #35 opened 2 years ago by Nehalalotaiby
  1

• 为什么不直接使用 insightface项目

  #17 opened 3 years ago by eeric
  11

• IJB-C Evaluation

  #34 opened 2 years ago by junwah712
  0

• Requirements problem

  #33 opened 2 years ago
  0

• visualize the data

  #31 opened 3 years ago by mtk380
  2

• GPU运行

  #32 opened 3 years ago by fenfen111
  2

• 请问一下数据增强

  #30 opened 3 years ago by supersyz
  1

• 请问能用自己的数据集进行训练吗，需要做什么处理吗？

  #29 opened 3 years ago by epns
  0

• Reproduce

  #28 opened 3 years ago by zws98
  6

• Latency and memory usage?

  #27 opened 3 years ago by johnnysclai
  1

• Could you add a LICENSE in the repo?

  #25 opened 3 years ago by zhoufengfan
  4

• how to solve 'size mismatch for patch_to_embedding.weight: copying a param with shape torch.Size([512, 432]) from checkpoint, the shape in current model is torch.Size([512, 192]).'?

  #23 opened 3 years ago by zhuyiliulian
  0

• 关于遮挡鲁棒性 About occlusion robustness

  #19 opened 3 years ago by ygtxr1997
  4

• 您好 最近在忙自己的论文 看到您发的这个论文感觉有很多值得借鉴的地方 想复现一下 想请问一下 图片切分成重叠的patch 这样做的目的是什么？

  #21 opened 3 years ago by zydongmingyu
  2

• 请问这个训练做了预处理吗？直接用的0-255做的输入吗？

  #20 opened 3 years ago by tianxingyzxq
  2

• Could you provide your traning log files?

  #18 opened 3 years ago by ygtxr1997
  2

• One GPU is working during Evaluation

  #15 opened 3 years ago by khawar-islam
  1

• Only feature extraction

  #16 opened 3 years ago by Aravinda27
  1

• Create .bin file for testing on more datasets

  #11 opened 4 years ago by khawar-islam
  1

• What's the difference between vit_face.py and vits_face.py？

  #10 opened 4 years ago by Gregory1994
  1

• NO gpu can run train.py?

  #9 opened 4 years ago by zouzouwei
  2

• Issues about input normalization

  #8 opened 4 years ago by zcy8123878
  2

• Have you tried patch size 16?

  #7 opened 4 years ago by szlbiubiubiu
  2

• Face Transformer Table I Query

  #6 opened 4 years ago by khawar-islam
  4

• Reproduce results using training

  #3 opened 4 years ago by khawar-islam
  4

• the acc of r100 trained by ms1m-retina

  #4 opened 4 years ago by xiangliu886
  0

• Data Accessibility problem

  #1 opened 4 years ago by khawar-islam
  2