Trends in Computer Vision with Amir Zamir - #338
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
00:00
Advancements in Self-Supervised Learning for Computer Vision
This chapter explores the evolution of self-supervised learning, highlighting its importance in computer vision by reducing dependence on labeled data. It examines key milestones from 2019, focusing on techniques like contrastive predictive coding that enhance image recognition capabilities. The discussion also addresses the relationship between self-supervised learning in NLP and computer vision, as well as advances in adversarial robustness of neural networks.
Transcript
Play full episode
