roatienza/deep-text-recognition-benchmark

Issues

• Demo.py

  #18 opened 3 years ago by Preethse
  3

• Predict `model.pth`

  #44 opened 2 months ago by minhduc01168
  2

• error in custom data

  #45 opened a month ago by MahmoudElsayedMahmoud
  1

• pretrained-model loading with errors

  #29 opened 2 years ago by Ao-Lee
  12

• Onnx conversion code trained model

  #43 opened 6 months ago by Shivashankarar
  0

• About Training time.

  #41 opened 7 months ago by DemissieDaniel
  0

• timm==0.4.5 ModuleNotFoundError: No module named 'timm.models.layers.patch_embed'

  #40 opened a year ago by sh1man
  1

• Train loss is 0.0000 at every iteration

  #38 opened a year ago by Ashish0091
  1

• About the speed of the model in Table 4 of the paper

  #37 opened 2 years ago by lexiaoyuan
  4

• Available Model weights.

  #36 opened 2 years ago by schreiterjp
  1

• ONNX

  #35 opened 2 years ago by centurions
  2

• CTC error

  #34 opened 2 years ago by LeeBronOff23
  1

• continue to train

  #33 opened 2 years ago by LeeBronOff23
  0

• train error

  #32 opened 2 years ago by chungluensing
  1

• Rand Aug

  #31 opened 2 years ago by fmobrj
  1

• quantification

  #30 opened 2 years ago by centurions
  0

• Training on Japanese data

  #10 opened 3 years ago by Preethse
  4

• about input channels of vitstr

  #28 opened 2 years ago by Name-Lessx
  1

• when i follow train.sh: line 6: --SequenceModeling: command not found

  #27 opened 2 years ago by felix115
  1

• About the parameter `--valid_data` in the training command mentioned in README.md

  #16 opened 3 years ago by lexiaoyuan
  1

• How to draw the attention map of ViTSTR?

  #24 opened 3 years ago by lexiaoyuan
  1

• Question about [GO] and [s]

  #26 opened 3 years ago by sparrow0629
  1

• How to calculate Top-5 accuracy?

  #25 opened 3 years ago by penghusile
  0

• Did you compare the result of CTC loss and cross entropy loss ?

  #17 opened 3 years ago by Jiakui
  0

• Training from scratch, w/o using Pretrained DeiT?

  #7 opened 3 years ago by mandal4
  3

• CUDA out of memory.

  #21 opened 3 years ago by penghusile
  1

• Is the network suit for long-text recognition?

  #6 opened 3 years ago by WudiJoey
  6

• Poor performance on some images

  #20 opened 3 years ago by dudeperf3ct
  1

• A question about [GO[ token

  #19 opened 3 years ago by zhaiyukun
  2

• About input size

  #15 opened 3 years ago by terryoo
  2

• why don't you normalize the images?

  #13 opened 3 years ago by cuongdxk57
  3

• Is there any performance comparison with clovaai/deep-text-recognition-benchmark

  #14 opened 3 years ago by LLC
  1

• model state loading issue

  #11 opened 3 years ago by rouarouatbi
  3

• I have a question

  #8 opened 3 years ago by daeing
  9

• about ACC

  #9 opened 3 years ago by YuNaruto
  2

• a question about ViTSTR

  #5 opened 3 years ago by Danee-wawawa
  2

• What is the meaning of 'delta' in plot_error.py? Is it the subtraction between previous model and ViTSTR-Base?

  #4 opened 3 years ago by huihui0000
  4

• About the difference between the number of training iters in the paper and this Repo

  #3 opened 3 years ago by superPangpang
  3

• AttributeError: 'NoneType' object has no attribute 'eval'

  #2 opened 3 years ago by momei123
  1

• Trained model?

  #1 opened 3 years ago by zobeirraisi
  1