Explore the cutting-edge research on modeling human visual object recognition mechanisms in this 50-minute lecture by James DiCarlo from MIT. Delve into topics such as encoding models, deep learning, and breakthroughs in the field. Gain insights into the big picture context and summary of visual object recognition research. Learn about neural population control, its goals, and potential long-term applications. Discover strategies for hypothesis building, organizing research, and forward engineering approaches for AI researchers. Engage with frequently asked questions and understand the practical applications of these models in advancing our understanding of human visual cognition.
The Quest to Create Engineer-Quality Models of the Mechanisms of Human Visual Object Recognition - Part 2
Overview
Syllabus
Intro
Encoding models
The big picture
Deep learning
Models of hype
Breakthrough
Big picture context
Big picture summary
Frequently asked questions
What can we do with these models
Neural population control
Population control goal
Negative control
Long term potential
Questions
Hypothesis Building
Organizing
Summary
Strategies for AI researchers
Forward engineering
Taught by
MITCBMM
