misf's People
Forkers
lincong666 githubcrj renhf316 zhumingxu cin-inpaiting hms97 masterhow hellpeo tayqiming xiaollz hzy-delmisf's Issues
关于网络的输出问题
请问网络中的输出是整张图片还是单纯mask擦除部分的预测结果?
我在阅读计算损失函数部分代码的时候觉得输出应该是整张图片,但是看代码在可视化的时候有一个计算outputs_merged的过程,并且在计算指标的时候也是通过outputs_merged和gt进行计算的。
Segmentation of celebA datasets
Regarding the division of the celebA dataset, in your experience, how many images are suitable for the test set and the validation set?
测试问题
你好,请问能加一下联系方式嘛,关于图像修复测试的问题,想向您请教一下。我的qq:2635505974
SFI-Swin: Symmetric Face Inpainting
Dear reaserchers, please also consider checking our newly introduced face inpainting method to address the symmetry problems of general inpainting methods by using swin transformer and semantic aware discriminators.
Our proposed method showed better results in terms of fid score and newly proposed metric which focus on the face symmetry compared to some of the state-of-the-art methods including lama.
Our paper is availabe at:
https://www.researchgate.net/publication/366984165_SFI-Swin_Symmetric_Face_Inpainting_with_Swin_Transformer_by_Distinctly_Learning_Face_Components_Distributions
The code also will be published in:
https://github.com/mohammadrezanaderi4/SFI-Swin
Test is not working
The test result is all black
log
Hello, may I ask if there is no log program
The outputs result is error
In the models.py, the process()
In the tracking results, it is found that the outputs result is a gray image.
What is the reason?
模型加载
您好,当测试的时候,预训练是怎么加载的呢?谢谢
训练中断可以从中断的地方开始训练吗?
训练次数太多
关于边缘
您好,在看您的代码时发现里面还保留有EdgeCon的 边缘预测模型,但实际上在您的代码中其实并不包含对边缘的使用对吗?也就是不需要先预测边缘再修补图像。
在训练中产生的数据会不会很大
作者您好,看了下在checkpoint中每一个epoch产生的文件就有100M左右,那几千个epoch所占用的空间非常大。还有train_sample和eval_sample,几万张图片训练的话产生的sample文件是不是也会非常大呀?
敦煌数据集无法修复
License
Thank you for sharing this work. Could you please upload a license file? Thanks
CelebA Dataset
Hi, can you provide the processed CelebA dataset as mentioned in your README.md file? The original CelebA dataset directory structure is different from yours. thank you
loss is nan
环境配置问题?
您好!我在运行时产生了多个与包版本相关的问题,版本冲突。您是否可以详细写下具体的包版本,比如python,pytorch,cuda。
网络不能处理奇数size的图像
一般全卷积网络可以处理任意大小的输入图片,但若输入图像size为奇数,则网络处理会有问题,请问应该如何修改?
the dunhuang dataset
hello,i can't the dataset of dunhuang, could you tell me about where to find this dataset??
Metrics
您好,首先您论文非常不错,最近我研究方向和您很相关。我想请问您的指标都是在256分辨率测试的吗,壁画有没有丢进模型去训练,我尝试运行壁画原分辨率,出现像素不一致问题。是否是您KPN模块出现问题。
谢谢,祝科研顺利!
Cannot test. Wrong dimensions
Hi!
Thanks for leaving here your project. Looks amazing! However, I wanted to test it and I got stucked. I downloaded the code and the models. Then, I use one of my images in jpg and after running the dataset I have two files, flist_real.txt and flist_mask.txt which contains:
["test_images\1.jpg"] and ["test_images\1_mask.jpg"] respectively. Now I ran the program and I got an error because of the dimensions. Should I have some special dimensions from the begginig? It's strange to see that if I use a png image instead of 6 channels error I got 8 channels...
What am I doing wrong? What version of pytorch and torchvision are you using?
Namespace(angle_aug=True, b1=0.5, b2=0.999, blind_est=True, burst_length=1, channel_att=False, color=True, core_bias=False, crop=False, crop_size=256, cudnn_benchmark=True, dataset=1, erode=19, eval_data='./data/face.txt', eval_interval=2, eval_mask='./data/mask_20.txt', eval_sample='./result/eval_sample', eval_sample_interval=1, geometry_aug=False, gpu_ids='0', init_gain=0.02, init_type='xavier', kernel_size=[3], load_name='./result/model/251000_KPN_bs_8_opt.pth', lr_decay=0.9, lr_decay_interval=1, lr_decrease_epoch=20, lr_g=0.0002, mask_threshold=100, mu=0, multi_gpu=False, num_workers=0, rainaug=False, save_mode_interval=1, save_model='./result/model/', scale_max=1, scale_min=1, sep_conv=False, sigma=30, spatial_att=False, test_batch_size=1, test_data='./data/dunhuang.txt', test_mask='./data/dunhuang_mask.txt', test_sample='./result/test_sample', test_sample_interval=1, train_batch_size=1, train_data='./data/face.txt', train_mask='./data/mask_20.txt', train_sample='./result/train_sample', train_sample_interval=1, upMode='bilinear', weight_decay=0)
initialize network with xavier type
Generator is created!
C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torchvision\models\_utils.py:209: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and will be removed in 0.15, please use 'weights' instead.
f"The parameter '{pretrained_param}' is deprecated since 0.13 and will be removed in 0.15, "
C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torchvision\models\_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=VGG19_Weights.IMAGENET1K_V1`. You can also use `weights=VGG19_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
Setting up [LPIPS] perceptual loss: trunk [vgg], v[0.1], spatial [off]
C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torchvision\models\_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=VGG16_Weights.IMAGENET1K_V1`. You can also use `weights=VGG16_Weights.DEFAULT` to get the most up-to-date weights.
warnings.warn(msg)
Loading model from: C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\lpips\weights\v0.1\vgg.pth
training:False mask:6 mask_list:./flist_mask.txt data_list:./flist_real.txt
test dataset:
Loading InpaintingModel generator...
start testing...
+++++++++++++++
Traceback (most recent call last):
File "test.py", line 2, in <module>
main(mode=2)
File "C:\Users\Scolymus\Desktop\misf\main.py", line 69, in main
model.test()
File "C:\Users\Scolymus\Desktop\misf\src\misf.py", line 395, in test
outputs = self.inpaint_model(images, edges, masks)
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\Scolymus\Desktop\misf\src\models.py", line 278, in forward
outputs = self.generator(inputs) # in: [rgb(3) + edge(1)]
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\Scolymus\Desktop\misf\src\networks.py", line 91, in forward
x = self.encoder0(x) # 64*256*256
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\container.py", line 139, in forward
input = module(input)
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\conv.py", line 457, in forward
return self._conv_forward(input, self.weight, self.bias)
File "C:\Users\Scolymus\.conda\envs\py37\lib\site-packages\torch\nn\modules\conv.py", line 454, in _conv_forward
self.padding, self.dilation, self.groups)
RuntimeError: Given groups=1, weight of size [64, 4, 7, 7], expected input[1, 6, 4006, 6006] to have 4 channels, but got 6 channels instead
训练问题
Program recurrence problem
Hello, your article is very good. I want to learn your model, but I get an error when I reproduce the program:
File "E:\LW2\misf\src\misf.py", line 36, in init
config.TRAIN_FLIST = tin3.tu_train
AttributeError: 'function' object has no attribute 'tu_train'
test error
dataset.py中img = Image.fromarray(img)这句是不是有问题
运行代码报错
File "D:\anaconda\envs\p37\lib\site-packages\PIL\Image.py", line 2992, in fromarray
mode, rawmode = _fromarray_typemap[typekey]
KeyError: ((1, 1, 3), '<f8')
改成img = Image.fromarray(np.uint8(img))后没有报错,但是图像数据有问题,都是0
训练的list文件
请问data_list.py生成的 .txt文件是真实图像和纯掩码图像,还是带有掩码的图像和纯掩码的图像呢
About requirements.txt !
Hi!
First of all, congratulations on your successful work.
I am interested in this work and want to run the testing code, but requirements.txt was uploaded empty.
So, can I get the environment information or requirements.txt(or yaml) about this code?
Thank you!
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.