Scattering Bricks to Build Invariants for Perception - Part 1

Explore a lecture on building invariants for perception through scattering bricks, presented by Stéphane Mallat from Ecole Normale Superieur. Delve into high-dimensional classification, addressing the curse of dimensionality and examining translations and deformations. Learn about stable translation invariants and the image wavelet transform. Discover wavelet translation invariance and techniques for recovering lost information. Investigate deep convolution networks and their scattering properties. Analyze linearized classification and the representation of random processes. Conclude by studying the classification of textures in this comprehensive exploration of advanced perception concepts.

Intro
High Dimensional Classification
Curse of Dimensionality
Translations and Deformations
Stable Translation Invariants
Image Wavelet Transform
Wavelet Translation Invariance
Recovering Lost Information
Deep Convolution Network
Scattering Properties
Linearized Classification
Representation of Random Processes
Classification of Textures