Scaling audio-visual learning without labels : comparemela.c

Researchers from MIT, the MIT-IBM Watson AI Lab, IBM Research, and elsewhere have developed a new technique for analyzing unlabeled audio and visual data that could improve the performance of machine-learning models used in applications like speech recognition and object detection. The work, for the first time, combines two architectures of self-supervised learning, contrastive learning and masked data modeling, in an effort to scale machine-learning tasks like event classification in sing...

Related Keywords

Frankfurt , Brandenburg , Germany , Texas , United States , Yuan Gong , Alexanderh Liu , Leonid Karlinsky , Hilde Kuehne , Andrew Rouditchenko , Artificial Intelligence Laboratory , International Conference On Learning Representations , Youtube , Goethe University Frankfurt , Ibm , University Of Texas At Austin , Watson Ai Lab , Bing Create , Computer Science , Jim Glass , David Harwath Phd , Goethe University , International Conference , Press Release Image , Researchers , It , It Ibm , Watson , Hi , Lab , Research , Eveloped , Technique , Analyzing , Unlabeled , Udio , Visual , Data , Mprove , Performance , Machine Learning , Models , Ased , Applications , Bike , Speech , Recognition , Object , Detection , Work , Dime , Combines , Architectures , Elf Supervised , Learning , Ontrastive , Asked , Modeling , Effort , Scale , Tasks , Event , Classification , King , Ebwire , Press Release , Ews Release ,