图像超分辨率ARCNN的Pytorch复现代码，注释详细，含科研绘图，各Quality下的最优SSIM和PSNR的模型权重文件

# AR-CNN, Fast AR-CNN This repository is implementation of the "Deep Convolution Networks for Compression Artifacts Reduction". <br /> In contrast with original paper, It use RGB channels instead of luminance channel in YCbCr space and smaller(16) batch size. ## Requirements - PyTorch - Tensorflow - tqdm - Numpy - Pillow **Tensorflow** is required for quickly fetching image in training phase. ## Results <table> <tr> <td><center>Input</center></td> <td><center>JPEG (Quality 10)</center></td> </tr> <tr> <td> <center><img src="./data/monarch.bmp" height="300"></center> </td> <td> <center><img src="./data/monarch_jpeg_q10.png" height="300"></center> </td> </tr> <tr> <td><center><b>AR-CNN</b></center></td> <td><center><b>Fast AR-CNN</b></center></td> </tr> <tr> <td> <center><img src="./data/monarch_ARCNN.png" height="300"></center> </td> <td> <center><img src="./data/monarch_FastARCNN.png" height="300"></center> </td> </tr> </table> ## Usages ### Train When training begins, the model weights will be saved every epoch. <br /> Data augmentation option **--use_augmentation** performs random rescale and rotation. <br /> If you want to train quickly, you should use **--use_fast_loader** option. ```bash python main.py --arch "ARCNN" \ # ARCNN, FastARCNN --images_dir "" \ --outputs_dir "" \ --jpeg_quality 10 \ --patch_size 24 \ --batch_size 16 \ --num_epochs 20 \ --lr 5e-4 \ --threads 8 \ --seed 123 \ --use_augmentation \ --use_fast_loader ``` ### Test Output results consist of image compressed with JPEG and image with artifacts reduced. ```bash python example --arch "ARCNN" \ # ARCNN, FastARCNN --weights_path "" \ --image_path "" \ --outputs_dir "" \ --jpeg_quality 10 ```