We invent audio AI technology for human-centered creativity. Our goal is to empower people to bring their creative ideas to life through high quality audio content. This ranges from music and podcasts to video and immersive experiences, as well as emerging media types. We are dramatically simplifying the creation process with AI so that people can quickly go from idea to produced content regardless of skill level, iterating on the creative aspects rather than the technical ones. We do this through our research in the analysis, processing, and generation of speech, music, everyday sounds, and more. We also work on problems at the intersection of audio with video, augmented reality, and natural language processing. To advance all of these research areas, we develop new machine learning models, novel signal processing algorithms, and new human computer interaction paradigms.
Project Awesome Audio, an experimental technology, converts low quality home voice recordings into studio quality audio at the press of a button. It uses a deep learning model to jointly perform high quality dereverberation, denoising, and more.