Sign In

视频动作迁移合集-图生视频-Video motion transfer-wan Animate、SteadyDancer、SCAIL

Type

Workflows

Stats

59

0

Reviews

Published

Feb 2, 2026

Base Model

Other

Hash

AutoV2
9EE6C9813C

Video Tutorial: https://www.youtube.com/watch?v=VuxvIlTw7Ho

Usage method

1、Model selection

Value [1] for use: wan Animate【Single character - Real human preferred】

Value [2] for use: SteadyDancer【Do not use for abnormal proportions cartoon】

Value [3] for use: Scail【Multiple people / Abnormal proportions cartoon / Complex action preferred】

2、Upload reference image【The number of people must be the same as in the reference video】

3、Upload reference action video

4、Fill in the video long side size【Default 1280, 640/960/1024/1280. The larger the value, the longer the generation time, and the better the video clarity】

Video long side size 1280, model【1】generation time about 2 minutes/second video, model【2】generation time about 2 minutes/second video, model【3】generation time about 3 minutes/second video. The generation time is based on the graphics card RTX4090

5、Fill in the number of characters in the reference video【Fill in according to the number of characters in the action reference video you uploaded】

6、Prompt words【Optional】

7、Model 1 facial reference (non-human needs to be turned off, such as: anime, animals, etc.) is only modified when using Model 1

It is recommended to try the online experience of this LORA before downloading, completely free!
Online experience address:https://www.runninghub.ai/post/2017059703095824385/?inviteCode=rh-v1058