EgoSonics: Generating Synchronized Audio for Silent Egocentric Videos

Aashish Rai, Srinath Sridhar

Brown University

[Project Page] [Arxiv]

We present EgoSonics, a method to synthesize audio tracks conditioned on silent in-the-wild egocentric videos. Our method operates on videos at 30 fps to synthesize audio that is semantically meaningful and synchronized with events in the video.

Code will be released upon acceptance.

Citation

If you find this paper useful, please consider citing:

@article{rai2024egosonics,
    title={EgoSonics: Generating Synchronized Audio for Silent Egocentric Videos},
    author={Rai, Aashish and Sridhar, Srinath},
    journal={arXiv preprint arXiv:2407.20592},
    year={2024}
}