Channel: Microsoft Research

Category: Science & Technology

Tags: visual learningbig data deep learningmicrosoft research summitreal-world tasksvisual taskshuman-like visual learningvisual reasoning

Description: Speakers: Suha Kwak, Professor, POSTECH Chong Luo, Principal Researcher, Microsoft Research Asia Lu Yuan, Principal Research Manager, Microsoft The limitations of big data-driven deep learning in scalability and adaptation to real-world scenarios hinder its practical applications. To address these limitations, it’s extremely important to develop architectures and algorithms that can capture the fundamentals of how humans learn, infer, and reason. Join Professor Suha Kwak from POSTECH, Microsoft Principal Researcher Chong Luo, and Microsoft Principal Research Manager Lu Yuan as they discuss the theory and practice of unsupervised visual representation learning for 3D point clouds, videos, and images, which help address model generalization and adaptation to new data domains and new downstream tasks. Learn more about the 2021 Microsoft Research Summit: Aka.ms/researchsummit