CodeFormer
CodeFormer copied to clipboard
[NeurIPS 2022] Towards Robust Blind Face Restoration with Codebook Lookup Transformer
My trained model is 700M?but yours is 300M?why?and how to do that?
when i am trying to finetune using pretrained weights, stage 3rd training, I am facing the below error. can anyone please help me? latent_gt = self.latent_gt_dict['orig'][name] [rank2]: KeyError: '000359' [rank3]:...
Hey! I have been trying to optimize the inference pipeline using batch inference. But everything used in the pipeline is written considering one image at a time. I was successful...
```yaml # general settings name: VQGAN-512-ds32-nearest-stage1 model_type: VQGANModel num_gpu: 4 manual_seed: 0 # dataset and data loader settings datasets: train: name: FFHQ type: FFHQBlindDataset dataroot_gt: /mnt/gc/dataset/FFHQ/images filename_tmpl: '{}' io_backend: type:...
Trying the **Codeformer Whole Image Enhancement** and getting amazing results but having some slight issue. It masks/crop the image into 512x512 everything inside will be restored. the face, including some...
我这电脑有amd显卡,可是运行的时候是cpu在跑,很慢,怎么改成amd的显卡
I am trying to implement Codeformer on the Android system. Are there any documents or repositories available for this? I would greatly appreciate any advice.
我昨天晚上用电脑下载了这个文件,早上在手机相册里发现了一张奇怪的照片,上面写AI换脸什么的,我不记得有保存过这张图片,请问我应该担心这个文件的安全性吗?
I found all the results are double eyelid, even the original one is single eyelid. 
