chenjun2hao/SRN.pytorch

Issues

• What is your license?

  #23 opened 2 years ago by frommoontosun
  1

• 没达到论文的acc是因为训练集?

  #20 opened 3 years ago by GaoXinJian-USTC
  0

• copying a param with shape torch.Size([38, 512]) from checkpoint, the shape in current model is torch.Size([39, 512]).

  #8 opened 4 years ago by robotSunGina
  4

• Alternative to download pre-trained model

  #9 opened 4 years ago by amrit-das
  3

• How to change module to fix more image size.

  #21 opened 3 years ago by maxh2010
  1

• acc is always 0

  #22 opened 3 years ago by wanglibing727
  0

• 你好，想问一下从头开始训练，数据集的格式应该是什么样的呢

  #19 opened 3 years ago by Henryplay
  0

• 关于论文中的attention map 可视化

  #17 opened 3 years ago by zdz1997
  2

• 用百度中文数据不收敛

  #16 opened 3 years ago by liudatutu
  0

• 论文中使用的TRW15测试集能上传下吗？数据集无法下载

  #18 opened 3 years ago by huihui0000
  0

• 和腾讯的2DAttentionalIrregularSceneTextRecognizer

  #4 opened 5 years ago by miaomi1994
  1

• pytorch模型转pt时，发生错误

  #15 opened 4 years ago by HaoRenkk123
  0

• File "SRN_modules.py", line 65, in forward return x + self.pos_table[:, :x.size(1)].clone().detach() RuntimeError: The size of tensor a (320) must match the size of tensor b (256) at non-singleton dimension 1

  #14 opened 4 years ago by wuxiaolianggit
  4

• why do we use two padding tokens ('$', and '#')

  #7 opened 4 years ago by yanfjz
  4

• 已有参考

  #10 opened 4 years ago by lanfengecho
  0

• GSRB中的argmax模块是否可导？ 应该不可导吧

  #12 opened 4 years ago by superPangpang
  1

• The PVAM module is different with paddleOcr

  #11 opened 4 years ago by archwolf118
  0

• 训练PAD问题

  #13 opened 4 years ago by wanlipeng
  0

• 没有复现论文的精度是否是训练集不同导致的？

  #3 opened 4 years ago by someoneAlready
  0

• why don't use resnet50FPN?

  #6 opened 5 years ago by miaomi1994
  0

• where can download the "BAIDU" datasets?

  #5 opened 5 years ago by chenyangMl
  0

• 默认 opt.input_channel = 1 ？

  #2 opened 5 years ago by someoneAlready
  1

• 关于中文文本识别率

  #1 opened 5 years ago by chibohe
  2