Voyager 3
Voyager 3
Here they are: [Google Drive](https://drive.google.com/drive/u/0/folders/17VYV_SoZZesU6mbxz2dMAIccSSlqLecY) and [Baidu Drive](https://pan.baidu.com/s/1-Lh6ma-wXzfH8NqeBtPaFQ) You can found them at [Readme](https://github.com/xinntao/ESRGAN) about 50% location
After reading the reference "_Image Blind Denoising With Generative Adversarial Network Based Noise Modeling_", I guess the Noise Extraction is like this ``` # in your dataloader's getitem() method: #...
DF2K-JPEG works really cool for me too 👍 @jixiaozhong hope to see the training code, because I use the kernalGAN and ESRGAN's training code, however I failed to achieve the...
Only magic 🧙 can beat magic 🧙 , so I guess adding color loss can address your problem 😃
I guess they need waiting for the patent, just keep patient
@yoon28 That's true. I implemented the paper using KernelGAN to estimate kernel and create a kernel pool, then using the ESRGAN to train paired images. However, my results is not...
@fuuuyuuu I try 3 kinds of Noise collection to create "noise pool", they are: 1. doble slide window in _Image Blind Denoising With Generative Adversarial Network Based Noise Modeling_ 2....
 the argorithm in the wgan-gp paper is above. There are two ways to calulate G and D iteratively in pytorch: 1. ``` # update D fake = G(z) fake.detach()...
>And SRFlow can't be trained on JPEGs or other artifacts. Why？
信创服务器都是arm64的，希望能尽快支持！