Bio

Ishan Misra is a Director, Research Scientist at Meta where he works on building machine learning models that can learn from large amounts of data using minimal supervision. His work spans the areas of multimodal learning, generative AI, and self-supervised learning. His work in self-supervised learning demonstrated how large computer vision models could perform better than supervised models. His work forms the basis of multiple techniques since 2019 and power AI models at Meta. He received the MIT Technology Review’s 35 innovators under 35 award for his contributions to self-supervised learning. Ishan holds a Ph.D. from Carnegie Mellon University where he worked with Martial Hebert and Abhinav Gupta. He also received Carnegie Mellon’s Recent Alumni Achievement Award for contributions to machine learning.