Explore deep generative models for brain reading in this 46-minute lecture by Mengmi Zhang from Harvard University and Children's Hospital, Boston. Delve into the intersection of deep learning and neuroscience, covering topics such as discriminator implementation in PyTorch, GAN training techniques, and methods for realistic image reconstruction from brain signals. Learn about evolving latent codes using genetic algorithms and mapping techniques with linear regression. Discover the potential applications of these technologies in brain-computer interfaces and gain insights into the latest advancements in decoding neural activity.
Toward Brain Computer Interface - Deep Generative Models for Brain Reading
Overview
Syllabus
Intro
Deep Learning in Machine Learning
Discriminator - pytorch implementation
Training GAN - Part 2
Tricks for More Realistic Image Reconstruction
Brain Reading
Image Reconstruction Methods from Brain Signals
Evolving Latent Code using Genetic Algorithm
Mapping Latent Code using Linear Regression
Mapping using Linear Regression
More brain scores - 15
Towards Brain Computer Interface
Taught by
MITCBMM
