← Back
Lex Friedman January 27, 2018 53m

MIT 6.S094: Computer Vision

Summary

The main theme is how machines "see" through computer vision, heavily relying on deep learning and neural networks. Key subjects include deep reinforcement learning, dynamic driving scene segmentation, and the crucial role of annotated data as ground truth for training. The practical takeaway is that understanding and effectively utilizing the raw data provided to machines is fundamental for successful computer vision tasks, even for experienced practitioners.

View original episode ↗