Bringing-Old-Photos-Back-to-Life

Bringing-Old-Photos-Back-to-Life copied to clipboard

Create VanceAI Photo Restorer

1

An online AI tool for [photo restoration](https://vanceai.com/old-photo-restoration/) allows you to eliminate scratches, fractures, and sepia within a few seconds

Tracey008

Sofia

2

SofiaVega

fix save model mapping

4

I don't know if this is a bug, but I can't save the model when training model mapping. I found that you didn't implement save function in `Pix2PixHDModel_Mapping`. I see...

PhanThanhTrung

can not save models for mapping_train

2

When I run train_mapping.py, the loss_log.txt and iter.txt are normal,but can not save models for mapping

wanna69

An online AI tool for **[old photo restoration](https://vanceai.com/old-photo-restoration/)** allows you to eliminate scratches, fractures, and sepia within a few seconds.

Tracey008

cp: cannot stat 'Synchronized-BatchNorm-PyTorch/sync_batchnorm': No such file or directory

2

cp -rf Synchronized-BatchNorm-PyTorch/sync_batchnorm . 在终端运行这一句时，发生以上错误，不太懂最后一个空格加个点是什么意思。

XiaoluJiayou

Global Restoration issue

2

After running: cd Global/ python test.py --Quality_restore \ --test_input [test_image_folder_path] \ --outputs_dir [output_path] on 1440x1080 input images, "output_dir" consists cropped 256x256 images

semel1

[enhancement] ONNX/Torchscript export

2

I have noticed `--export_onnx` flag in `test_options.py`. It would be nice to have onnx/torchscript conversion.

p1x31

Cannot cast ufunc 'multiply' output from dtype('float64') to dtype('uint8') with casting rule 'same_kind'

4

Running run.py with parameters --with_scratch --HR --GPU 0 on Windows 11, Python 3.9, numpy 1.22.4 gives Traceback (most recent call last): File "D:\Users\theke\python projects\BringingOldPicturesBackToLife\Bringing-Old-Photos-Back-to-Life\Face_Detection\align_warp_back_multiple_dlib_HR.py", line 428, in blended = blur_blending_cv2(warped_back,...

pyrator

Face_Enhancement model converting to onnx failure

7

Hello, Thanks for the great work. I'm trying to convert this model into onnx, but have met a few issues. The `mv` and `dot` operator used by PyTorch's `spectral_norm` was...

cjenkins5614