Fully Automatic Perfectly Copy Actions and Music

Can run 150 frames currently, non-frame interpolation single generation, the highest across all platforms
The setting here is 5 seconds, 30 frames/second

2025/6/19 Update HD output 1920*1080
2025/7/1 Update Enhanced facial description (choose whether to include glasses)

2025/7/7 Update
Dance duration (seconds)
Skip duration (seconds)
Frame rate (if you fill in 16, you can try 10s per generation)
Understanding intensity (optimize consistency, but may cause image crashes)

Operation Steps
1. Insert your favorite character image.
2. Insert a reference video.

The following are points to note, which can help you generate better results.
1. Character image reference sample. Do not make it compact, leave some space around the character
2. Reference videos are recommended to focus on the upper body, clearly showing body movements for better capture and positioning

The tested single time is 15-22 minutes. Be patient after input.

Detailed operations can use the workflow version. Faster speed
Reducing the frame count can improve other performance indicators
Individual video processing
150 frames. Generated video takes 8-13 minutes per time
Individual action processing
150 frames. Action processing takes 6-7 minutes per time