Michael Rubinstein - Automatically Isolating Individual Voices in Noisy Videos
Keynote: Automatically Isolating Individual Voices in Noisy Videos
Videos frequently have people speaking simultaneously over each other or speaking in noisy rooms, which makes it difficult to hear individual speakers. Michael will inspire with his latest work leveraging both audio and video for isolating a single speech signal from a mixture of sounds in noisy environments. Valuable use cases for this solution range from enhancing videos, speech recognition, captioning, video conferencing, and solutions across the Internet of Eyes such as smart assistants. Michael is a Senior Research Scientist at Google, received his PhD from MIT under the supervision of Bill Freeman and spent a year as a postdoc at Microsoft Research New England.
- Michael Rubinstein, Google, Research Scientist