Efthymios Tzinis
Efthymios (Thymios) is a research scientist working on uni- and multi-modal sound separation and sound understanding in general. His research interests include the development of self-supervised audio and video perception machine learning models and their efficient deployment on low-energy mobile devices. He got his Ph.D. in Computer Science from the University of Illinois at Urbana-Champaign and also holds a diploma (B.Sc. + M.Eng.) in Electrical and Computer Engineering from the National Technical University of Athens. He has co-authored more than 30 papers in the field of audio and audio-visual processing and holds one US patent. Efthymios has also served as a reviewer to most of the top-notch journals and annual international conference-venues in the fields of deep learning and audio-visual processing and has also been recognized as an outstanding reviewer award 4 times. He has also been a past recipient of the Google Ph.D. fellowship for Machine Perception, Speech Technology and Computer Vision.
Research Areas
Authored Publications
Sort By
Google
Don’t Listen to What You Can’t See: The Importance of Negative Examples for Audio-Visual On-Screen Sound Separation
ECCV 2022 Workshop on AV4D: Visual Learning of Sounds in Spaces
AudioScopeV2: Audio-Visual Attention Architectures for Calibrated Open-Domain On-Screen Sound Separation
Tal Remez
European Conference on Computer Vision (ECCV) (2022)
Into the Wild with AudioScope: Unsupervised Audio-Visual Separation of On-Screen Sounds
Tal Remez
International Conference on Learning Representations (ICLR) 2021
Self-Supervised Audio-Visual Separation of On-Screen Sounds from Unlabeled Videos
Tal Remez
NeurIPS 2020 Workshop on Self-Supervised Learning for Speech and Audio Processing
Unsupervised Speech Separation Using Mixtures of Mixtures
ICML 2020 Workshop on Self-Supervision for Audio and Speech
