基于AdvancedLivePortrait针对DeepFaceLab的src素材角度不足问题改版

WaveBedo

冬日可爱~ 发表于 2025-6-12 18:59
确实挺早的，我之前没逛论坛（主要是没灵石），b站看到个类似的还收费。就自己动手改了一下，果然总会有 ...

还有一个坛友更早，再早一个月，他是使用comfyui做的
liveportrait批量修改
https://dfldata.cc/forum.php?mod=viewthread&tid=19968

如果不是他的角度比较小，20度，我还是挺想用他这个版本的
他这个版本有一个功能，就是例如，你要补充闭眼的素材，
你可以先拿一张睁眼的图片，先调整参数，让他闭眼，然后再按批量
他这个可以一步到位

如果是其他版本的话，就需要两步才能做得到。
多一步的话，应该会增加模糊的几率。
这个我是猜测的，没有对比过

test001

5060ti 报错  顺便可否告知修改角度位置？
raceback (most recent call last):
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\gradio\queueing.py", line 625, in process_events
    response = await route_utils.call_process_api(
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\gradio\route_utils.py", line 322, in call_process_api
    output = await app.get_blocks().process_api(
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\gradio\blocks.py", line 2218, in process_api
    result = await self.call_function(
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\gradio\blocks.py", line 1729, in call_function
    prediction = await anyio.to_thread.run_sync(  # type: ignore
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\anyio\to_thread.py", line 56, in run_sync
    return await get_async_backend().run_sync_in_worker_thread(
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\anyio\_backends\_asyncio.py", line 2470, in run_sync_in_worker_thread
    return await future
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\anyio\_backends\_asyncio.py", line 967, in run
    result = context.run(func, *args)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\gradio\utils.py", line 894, in wrapper
    response = f(*args, **kwargs)
  File "E:\AdvancedLivePortrait-WebUI\app\modules\live_portrait\live_portrait_inferencer.py", line 184, in edit_expression
    self.psi = self.prepare_source(src_image, crop_factor)
  File "E:\AdvancedLivePortrait-WebUI\app\modules\live_portrait\live_portrait_inferencer.py", line 586, in prepare_source
    crop_region = self.detect_face(img_rgb, crop_factor)
  File "E:\AdvancedLivePortrait-WebUI\app\modules\live_portrait\live_portrait_inferencer.py", line 434, in detect_face
    bboxes = self.get_face_bboxes(image_rgb)
  File "E:\AdvancedLivePortrait-WebUI\app\modules\live_portrait\live_portrait_inferencer.py", line 427, in get_face_bboxes
    pred = self.detect_model(image_rgb, conf=0.7, device=self.device)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\engine\model.py", line 179, in __call__
    return self.predict(source, stream, **kwargs)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\engine\model.py", line 550, in predict
    self.predictor.setup_model(model=self.model, verbose=is_cli)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\engine\predictor.py", line 308, in setup_model
    self.model = AutoBackend(
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\nn\autobackend.py", line 148, in __init__
    model = model.fuse(verbose=verbose)
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\nn\tasks.py", line 207, in fuse
    m.conv = fuse_conv_and_bn(m.conv, m.bn)  # update conv
  File "E:\AdvancedLivePortrait-WebUI\python\lib\site-packages\ultralytics\utils\torch_utils.py", line 261, in fuse_conv_and_bn
    w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var)))
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

冬日可爱~

WaveBedo 发表于 2025-6-13 00:58
还有一个坛友更早，再早一个月，他是使用comfyui做的
liveportrait批量修改
https://dfldata.cc/forum.ph ...

我这个角度只是个建议值，你可以输入填100都行，就是滑块没有那么大

冬日可爱~

test001 发表于 2025-6-13 04:53
5060ti 报错  顺便可否告知修改角度位置？
raceback (most recent call last):
  File "E:\AdvancedLivePor ...

没有50系显卡，这个项目的版本似乎还没适配，所以这个我也不知道怎么搞

冬日可爱~

WaveBedo 发表于 2025-6-13 00:58
还有一个坛友更早，再早一个月，他是使用comfyui做的
liveportrait批量修改
https://dfldata.cc/forum.ph ...

我这个随便哪个参数也都可以单独批量生成的，只调需要的参数就行了，也算是一步到位吧。