memoavatar/memo
A diffusion model that generates expressive talking videos from audio and a reference face image.
Not currently ranked — collecting fresh signals.
star history
MEMO is a memory-guided diffusion model for synthesizing realistic talking head videos with natural facial expressions and head movements. It takes audio input and a reference face image to generate temporally consistent video sequences. The model is available on Hugging Face with community integrations including ComfyUI and Gradio apps.
Frequently asked
- What is memoavatar/memo?
- A diffusion model that generates expressive talking videos from audio and a reference face image.
- Is memo open source?
- Yes — memoavatar/memo is open source, released under the Apache-2.0 license.
- What language is memo written in?
- memoavatar/memo is primarily written in Python.
- How popular is memo?
- memoavatar/memo has 1.1k stars on GitHub.
- Where can I find memo?
- memoavatar/memo is on GitHub at https://github.com/memoavatar/memo.