search

RQ-Wu / LAMP

Official implement code of LAMP: Learn a Motion Pattern by Few-Shot Tuning a Text-to-Image Diffusion Model (Few-shot-based text-to-video diffusion)
https://rq-wu.github.io/projects/LAMP/index.html
Other
236 stars 10 forks source link

About multi-action #14

Open fradino opened 6 months ago

fradino commented 6 months ago

Thank you for your excellent work! I try to train three actions at the same time(horse run, birds fly, waterfall), but the result is not as good as the single action. Can you give me some suggestions?

RQ-Wu commented 6 months ago

Learning multiple actions on limited data is difficult. In my experiments, learning a single action can also cause performance degradation if the given data is very different. Some continuous learning ideas (e.g. reparameterization) will help you. It makes sense to solve this problem.

ps: If you feel our project is useful, please give us a little star🌟 :)