How to use the trained model?

[nissansz]

No description provided.

[twelfth-star]

If the trained CNN model is saved as ../data/models/cnn_model.torch, then you can use the following code

import torch

cnn_net = torch.load('../data/models/cnn_model.torch')

img = Image.open(img_path)
img = img.copy().convert("L")

label = cnn_net(img)

The label will be the ID of the font.

[nissansz]

Do you have pre-trained model?
How to use onnx version for such model?

I tried https://github.com/HemingwayLee/deepfont-implement, and the result seems not correct for Japanese fonts.

[twelfth-star]

I have trained a toy model for Japanese fonts months ago. You can download the model here. It can only recognize the following five fonts. You can download the fonts here.

• GenEiAntiqueNv5-M.ttf
• GenEiKoburiMin6-R.ttf
• GenEiLateMinN_v2.ttf
• GenEiMGothic2-Bold.ttf
• GenEiMGothic2-Regular.ttf

I have refactored my code and I am currently trying to train a larger model which can recognize 169 Japanese fonts. I may release it soon.

[nissansz]

Good. Will wait for your large model to test.
Ms gothic etc.

[nissansz]

import torch

cnn_net = torch.load('../data/models/cnn_model.torch')

from PIL import Image

cnn_net = torch.load(r'L:/font/cnn_model.torch')

img = Image.open(img_path)

img = Image.open(r'F:/24.png')
img = img.copy().convert("L")

label = cnn_net(img)
print(label)

报错===========================================

D:\Python3.8.8\python.exe F:/pycharm2020.2/universal-font-recognition-main/eval.py
Traceback (most recent call last):
File "F:/pycharm2020.2/universal-font-recognition-main/eval.py", line 12, in
label = cnn_net(img)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\module.py", line 1194, in _call_impl
return forward_call(*input, **kwargs)
File "F:\pycharm2020.2\universal-font-recognition-main\utils\CNN.py", line 48, in forward
X = self.Cu(X)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\module.py", line 1194, in _call_impl
return forward_call(*input, **kwargs)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\container.py", line 204, in forward
input = module(input)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\module.py", line 1194, in _call_impl
return forward_call(*input, **kwargs)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\conv.py", line 463, in forward
return self._conv_forward(input, self.weight, self.bias)
File "D:\Python3.8.8\lib\site-packages\torch\nn\modules\conv.py", line 459, in _conv_forward
return F.conv2d(input, weight, bias, self.stride,
TypeError: conv2d() received an invalid combination of arguments - got (Image, Parameter, Parameter, tuple, tuple, tuple, int), but expected one of:

• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, tuple of ints padding, tuple of ints dilation, int groups)
  didn't match because some of the arguments have invalid types: (!Image!, !Parameter!, !Parameter!, !tuple!, !tuple!, !tuple!, int)
• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, str padding, tuple of ints dilation, int groups)
  didn't match because some of the arguments have invalid types: (!Image!, !Parameter!, !Parameter!, !tuple!, !tuple!, !tuple!, int)

Process finished with exit code 1