Character precise motion control (test), improved on the basis of Animatediff SparseCtrl img2img2vid 4/8frames2VideoPrediction LLM SD15, adding key frame motion control, attention mask, combining SparseCtrl with the latest version of Ip-adapter-plus plus ip_plus_composition_sd15, which can be used for single img generates 16/24/32 frame video. The lower the frame rate, the more stable it is. The higher the frame rate, the greater the range of motion.
Frontal, side, clear posture , and img with a larger proportion of faces have better results.
PS: There is still a problem with video upscale. It consumes a lot of video memory and runs slowly, so it is temporarily unavailable.
人物精准动作控制(测试),在Animatediff SparseCtrl img2img2vid 4/8frames2VideoPrediction LLM SD15基础上改进,增加了关键帧动作控制,注意力遮罩,将SparseCtrl与最新版Ip-adapter-plus加ip_plus_composition_sd15结合,可以单张图生成16/24/32帧视频,帧率越低越稳定,帧率越高动作幅度越大。
正面,侧面,人脸占比较大的图,效果较好。
PS:视频放大还有问题, 消耗大量显存运行缓慢,暂时不能用。
Description
FAQ
Comments (7)
Getting this error message in the first KSampler (Efficient) node:
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!
Any thoughts on what could be causing this?
Please set GPU as the only option in the comfyui-manager, or make sure the text format in BatchPromptSchedule meets the requirements, and there should be no extra symbols, especially the last line.
btw,the last line in BatchPromptSchedule-text, make sure don't enter too many prompts
@1984batman Forgive me, but I can't find any option in the comfyui-manager where I can set GPU or CPU. Would you describe exactly where I might find this option?
Also, what requirements are there for the BatchPromptSchedule? I just left it as you posted it here - which is blank.
@97Buckeye I have not encountered an error message similar to yours, but the vector/tensors-related errors that occur in the KSample node are probably related to the text in the BatchPromptSchedule node (KeyFramesAction in the workflow). This node has strict syntax format requirements. In addition, it is recommended to google.
Hi,
I cant get it run, could you explain more in workflow exact models u use with links?
OK
