Rope 换脸直播版 V6版 - 支持50系显卡 视频换脸/实时直播换脸软件 本地一键整合包下载
最近很火的一个直播实时换脸软件，基于 Roop珍珠版 ，支持一张照片换脸及实时直播换脸，相比较其他傻瓜实时换脸软件，Rope直播版实时换脸效果更好，实时换脸速度更加流畅，以及资源占用更低，6G显存就能流畅运行。
Rope直播版支持本地摄像头、虚拟摄像头、以及OBS一键发送到推流软件比如直播伴侣中开启直播换脸。无需单独捕获窗口，画质更清晰、流畅。支持TensorRT加速，实时换脸资源占用更低，速度更快。支持人脸编辑，比如调节五官和表情。
今天分享的Rope 直播版因是本站发布的第六个版本，故名V6版，主要新增了对50系显卡的支持（50系显卡下CUDA使用正常，TensorRT有问题，请看清楚再下载）。
使用教程：(只支持N卡，显存6G起。建议50系显卡使用，非50系使用 V5版)
下载链接内附视频教程
下载地址：（下载前，请先阅读 使用说明 再下载，避免不必要的问题和纠纷）
夸克网盘：**** 本内容需购买 ****
百度网盘：
**** 本内容需购买 ****
解压密码：https://deepface.cc/ 复制这个完整的网址即是解压密码，不要有空格，复制粘贴即可
不错，非常喜欢！！！ 首先是要支持，其次就是宝藏资源太多了，必须要不断地学习 求助，5070检测不出人脸 报错：
Microsoft Windows [版本 10.0.26100.4349]
(c) Microsoft Corporation。保留所有权利。
D:\AIYTHL-50Series>deepface\python main.py
Custom GUI.py: Loading and applying modifications (v2)...
Custom GUI.py: Original GUI module loaded from D:\AIYTHL-50Series\rope\GUI.pyc.
Custom GUI.py: Script finished. Custom GUI class is now exposed.
CustomGUI init called.
CustomGUI: Original GUI init finished.
CustomGUI: Applying early customizations (title & icon)...
CustomGUI: Window title changed in early customizations.
CustomGUI: Calculated custom icon path for early customizations: D:\AIYTHL-50Series\123.ico
CustomGUI: Custom icon file 'D:\AIYTHL-50Series\123.ico' NOT FOUND for early customizations.
CustomGUI: Applying UI modifications in create_gui (labels)...
CustomGUI: UI modifications (labels) applied successfully in create_gui.
[ WARN:1@11.981] global cap_msmf.cpp:476 anonymous-namespace'::SourceReaderCB::OnReadSample videoio(MSMF): OnReadSample() is called with error status: -1072875772 [ WARN:1@11.988] global cap_msmf.cpp:488 anonymous-namespace'::SourceReaderCB::OnReadSample videoio(MSMF): async ReadSample() call is failed with error status: -1072875772
[ WARN:0@11.995] global cap_msmf.cpp:1795 CvCapture_MSMF::grabFrame videoio(MSMF): can't grab frame. Error: -1072875772
global obsensor_uvc_stream_channel.cpp:158 cv::obsensor::getStreamChannelGroup Camera index out of range
Exception in Tkinter callback
Traceback (most recent call last):
File "D:\AIYTHL-50Series\deepface\Lib\tkinter_init_.py", line 1967, in call
return self.func(*args)
^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\GUIElements.py", line 763, in <lambda>
self.button.configure(command=lambda: self.function())
^^^^^^^^^^^^^^^
File "GUI.py", line 1960, in load_all
File "GUI.py", line 2134, in load_input_faces
File "D:\AIYTHL-50Series\rope\Models.py", line 197, in run_detect
bboxes, kpss_5, kpss = self.detect_retinaface(img, max_num=max_num, score=score, use_landmark_detection=use_landmark_detection, landmark_detect_mode=landmark_detect_mode, landmark_score=landmark_score, from_points=from_points, rotation_angles=rotation_angles)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\Models.py", line 1047, in detect_retinaface
self.retinaface_model.run_with_iobinding(io_binding)
File "D:\AIYTHL-50Series\deepface\Lib\site-packages\onnxruntime\capi\onnxruntime_inference_collection.py", line 381, in run_with_iobinding
self._sess.run_with_iobinding(iobinding.iobinding, run_options)
RuntimeError: Error in execution: Non-zero status code returned while running TRTKernel_graph_torch-jit-export_18153684661485309312_0 node. Name:'TensorrtExecutionProvider_TRTKernel_graph_torch-jit-export_18153684661485309312_0_0' Status Message: TensorRT EP failed to create engine from network.
Exception in Tkinter callback
Traceback (most recent call last):
File "D:\AIYTHL-50Series\deepface\Lib\tkinter_init.py", line 1967, in call
return self.func(*args)
^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\GUIElements.py", line 763, in <lambda>
self.button.configure(command=lambda: self.function())
^^^^^^^^^^^^^^^
File "GUI.py", line 2000, in select_faces_path
File "GUI.py", line 2134, in load_input_faces
File "D:\AIYTHL-50Series\rope\Models.py", line 197, in run_detect
bboxes, kpss_5, kpss = self.detect_retinaface(img, max_num=max_num, score=score, use_landmark_detection=use_landmark_detection, landmark_detect_mode=landmark_detect_mode, landmark_score=landmark_score, from_points=from_points, rotation_angles=rotation_angles)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\Models.py", line 1047, in detect_retinaface
self.retinaface_model.run_with_iobinding(io_binding)
File "D:\AIYTHL-50Series\deepface\Lib\site-packages\onnxruntime\capi\onnxruntime_inference_collection.py", line 381, in run_with_iobinding
self._sess.run_with_iobinding(iobinding.iobinding, run_options)
RuntimeError: Error in execution: Non-zero status code returned while running TRTKernel_graph_torch-jit-export_18153684661485309312_0 node. Name:'TensorrtExecutionProvider_TRTKernel_graph_torch-jit-export_18153684661485309312_0_0' Status Message: No engine is found.
[ WARN:2@100.697] global cap_msmf.cpp:476 anonymous-namespace'::SourceReaderCB::OnReadSample videoio(MSMF): OnReadSample() is called with error status: -1072875772 [ WARN:2@100.712] global cap_msmf.cpp:488 anonymous-namespace'::SourceReaderCB::OnReadSample videoio(MSMF): async ReadSample() call is failed with error status: -1072875772
[ WARN:0@100.730] global cap_msmf.cpp:1795 CvCapture_MSMF::grabFrame videoio(MSMF): can't grab frame. Error: -1072875772
global obsensor_uvc_stream_channel.cpp:158 cv::obsensor::getStreamChannelGroup Camera index out of range
Exception in Tkinter callback
Traceback (most recent call last):
File "D:\AIYTHL-50Series\deepface\Lib\tkinter_init.py", line 1967, in call
return self.func(*args)
^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\GUIElements.py", line 763, in <lambda>
self.button.configure(command=lambda: self.function())
^^^^^^^^^^^^^^^
File "GUI.py", line 1960, in load_all
File "GUI.py", line 2134, in load_input_faces
File "D:\AIYTHL-50Series\rope\Models.py", line 197, in run_detect
bboxes, kpss_5, kpss = self.detect_retinaface(img, max_num=max_num, score=score, use_landmark_detection=use_landmark_detection, landmark_detect_mode=landmark_detect_mode, landmark_score=landmark_score, from_points=from_points, rotation_angles=rotation_angles)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "D:\AIYTHL-50Series\rope\Models.py", line 1047, in detect_retinaface
self.retinaface_model.run_with_iobinding(io_binding)
File "D:\AIYTHL-50Series\deepface\Lib\site-packages\onnxruntime\capi\onnxruntime_inference_collection.py", line 381, in run_with_iobinding
self._sess.run_with_iobinding(iobinding._iobinding, run_options)
RuntimeError: Error in execution: Non-zero status code returned while running TRTKernel_graph_torch-jit-export_18153684661485309312_0 node. Name:'TensorrtExecutionProvider_TRTKernel_graph_torch-jit-export_18153684661485309312_0_0' Status Message: No engine is found.
taiyangbaobei 发表于 2025-6-19 11:12
求助，5070检测不出人脸 报错：
Microsoft Windows [版本 10.0.26100.4349]
(c) Microsoft Corporation。保 ...
常见注意事项
人脸不要太大，不要用大头照，也不要太小太模糊，最好用露出肩膀那种，图像存放路径，文件名等不要有中文。图片格式最好标准jpg格式。 无言以对 发表于 2025-6-19 11:29
常见注意事项
人脸不要太大，不要用大头照，也不要太小太模糊，最好用露出肩膀那种，图像存放路径，文件 ...
这些都没问题，使用CUDA模式可以运行，但太卡，想用Tensorrt加速，但这时就会出现上述提示 taiyangbaobei 发表于 2025-6-19 11:34
这些都没问题，使用CUDA模式可以运行，但太卡，想用Tensorrt加速，但这时就会出现上述提示 ...
5070这么拉跨吗？我用8G的5060测试了这个版本，CUDA模式下都很流畅。这个版本TensorRT可能有问题，因为Rope停更了，所以主要精力放VisoMaster了，VisoMaster就是Rope的更新版
可以用VisoMaster V6版，这个支持50系的TensorRT模式
本帖最后由 taiyangbaobei 于 2025-6-19 12:21 编辑
无言以对 发表于 2025-6-19 11:45
5070这么拉跨吗？我用8G的5060测试了这个版本，CUDA模式下都很流畅。这个版本TensorRT可能有问题，因为Rop ...
OKOKOKOKOKOK 使用本地摄像头实时换脸的时候，手摸脸的时候会变形，怎么解决
实时换脸以后，将帧发送到虚拟摄像头。一开始可以使用。后面电脑重新启动以后。就不能使用，虚拟摄像头，没有显示内容。