【人脸图清晰化神器】GFPGAN_GPEN_CodeFormer 3合1人像增强工具集

EineKleine

感谢分享

by4

不错不错

renhekeli

感谢分享

renhekeli

用2048模型出现错误
Traceback (most recent call last):
  File "run.py", line 170, in <module>
    img_out, orig_faces, enhanced_faces = processer.process(img, aligned=args.aligned)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_enhancement.py", line 83, in process
    ef = self.facegan.process(of)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\face_gan.py", line 42, in process
    out, __ = self.model(img_t)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\_internal\Python37\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\gpen_model.py", line 689, in forward
    outs = self.generator([outs], return_latents, inject_index, truncation, truncation_latent, input_is_latent, noise=noise[1:])
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\_internal\Python37\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\gpen_model.py", line 544, in forward
    out = conv1(out, latent[:, i], noise=noise1)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\_internal\Python37\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\gpen_model.py", line 354, in forward
    out = self.activate(out)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\_internal\Python37\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\op\fused_act.py", line 89, in forward
    return fused_leaky_relu(input, self.bias, self.negative_slope, self.scale, self.device)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\GPEN\face_model\op\fused_act.py", line 96, in fused_leaky_relu
    return scale * F.leaky_relu(input + bias.view((1, -1)+(1,)*(len(input.shape)-2)), negative_slope=negative_slope)
  File "D:\1111\GPEN_GFPGAN220314\GPEN_GFPGAN_Windows\_internal\Python37\lib\site-packages\torch\nn\functional.py", line 1474, in leaky_relu
    result = torch._C._nn.leaky_relu(input, negative_slope)
RuntimeError: CUDA out of memory. Tried to allocate 512.00 MiB (GPU 0; 4.00 GiB total capacity; 2.29 GiB already allocated; 0 bytes free; 2.35 GiB reserved in total by PyTorch)

lucq3000

大佬厉害！

jackliu

为啥缺少1024模型呢？

renzz

图形界面的，感谢，

leeso

renhekeli 发表于 2022-3-15 20:52
用2048模型出现错误
Traceback (most recent call last):
  File "run.py", line 170, in

显存不足  2048不适合你

sdfhj

请问有没有推荐的包括图片中人物背景一起修复增强的开源项目？

joezhangyn

最新好像两个都有更新版了