We are re-imagining audio technology to empower people to be creative without the tools getting in the way. We are working on dramatically simplifying the creation of audio content, while still maintaining high production value. Our work spans a number of audio research areas including speech enhancement, music information retrieval, speech and music synthesis, computational acoustics, spatial audio, and audio event detection. 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 algorithms, novel signal processing algorithms, and new human computer interaction paradigms.
SoundSeek, an experimental technology, allows users to find any sound in an audio track quickly and easily. All the user has to do is select one or more examples of the target sound, and SoundSeek will find everywhere else in the recording where a similar sound occurs, using a few-shot deep learning model.