This repository contains educational resources that I've used for mentoring students in introductory computational and visual neuroscience. It features a study plan, targeted questions, and hands-on projects.
Below are some resources that have proven valuable in my own journey. I am a firm believer that complexity in learning new subjects often comes from the way the material is presented rather than the content itself. Therefore, I've selected these resources, making sure they are not only informative but also accessible.
- Neuronal Dynamics - Computational Neuroscience by Prof. Wulfram Gerstner et al.. This Lecture series offers an excellent overview of the computational aspect of neuroscience. Previously available on edX, it's now freely accessible on YouTube.
- Vision and Brain: How We Perceive the World by Dr. James V. Stone. This is a fantastic book that provides technical depth yet remains enjoyable to read. You can find the book here. We'll cover the first six chapters, which align with our focus on visual neuroscience.
- Principles of Neural Information Theory: Computational Neuroscience and Metabolic Efficiency also by Dr. James V. Stone: This book beautifully applies information theory to neuroscience. We'll be reading Chapters 2 through 8. You can find the book here.
To complement your reading, I've converted my own notes into a set of questions for each week. These are designed to be more of a study companion rather than just an additional task.
To help you gain practical experience, I've included projects involving spiking neural networks, using the Nengo platform. Nengo, developed by Dr. Chris Eliasmith and his team at the University of Waterloo, is a well-supported tool with user-friendly interface (shown above, figure from their webpage) and comprehensive tutorials. Throughout the course, you'll get comfortable using Nengo and by the end, you'll be creating your own model.
Abbreviations of reading materials:
- ND = Neuronal Dynamics - Computational Neuroscience lecture videos
- V&B = Vision and Brain: How We Perceive the World
- PNIT = Principles of Neural Information Theory: Computational Neuroscience and Metabolic Efficiency
| Week | Readings | Assignments |
|---|---|---|
| 1 | • ND: Lecture CNS1 - A First Simple Neuron Model • V&B: Chapter 1 - Vision: An Overview |
• Question Set 1 • Nengo: Installation and Single Neuron Model |
| 2 | • ND: Lecture CNS2 - The Hodgkin-Huxley and Ion-Current Models • V&B: Chapter 2 - Eye |
• Question Set 2 • Nengo: Scalar and Vector |
| 3 | • ND: Lecture CNS3 - Synapses, Dendrites and the Cable Equation • V&B: Chapter 3 - The Neuronal Machinery of Vision |
• Question Set 3 • Nengo: Addition and Linear Transform |
| 4 | • ND: Lecture CNS4 - Two-dimensional models and phase plane analysis • V&B: Chapter 4 - The Visual Brain |
• Question Set 4 • Nengo: Nonlinear Transform |
| 5 | • ND: Lecture CNS5 - Variability of spikes trains • V&B: Chapter 5 - Depth: The Rogue Dimension |
• Question Set 5 • Nengo: Structured Representations |
| 6 | • ND: Lecture CNS6 - Noise models • V&B: Chapter 6 - The Perfect Guessing Machine |
• Question Set 6 • Nengo: Question Answering |
| 7 | • ND: Lecture CNS7 - Modern phenomenological neuron models • PNIT: Chapter 2 - Information Theory |
• Question Set 7 • |
| 8 | • ND: Lecture NDC3 - Neuronal Populations • PNIT: Chapter 3 - Measuring Neural Information |
• Question Set 8 • |
| 9 | • ND: Lecture CNS8 Fokker-Planck equation for stochastic integrate-and-fire neurons • PNIT: Chapter 4 - Pricing Neural Information |
• Question Set 9 • |
| 10 | • ND: Lecture NDC1 - Associative Memory in a Network of Neurons • PNIT: Chapter 5 - Encoding Colour |
• Question Set 10 • |
| 11 | • ND: Lecture NDC2 Attractor Networks and Generalizations of the Hopfield model • PNIT: Chapter 6 - Encoding Time |
• Question Set 11 • |
| 12 | • ND: Lecture NDC4 - Continuum models: Cortical Fields and Perception • PNIT: Chapter 7 - Encoding Space |
• Question Set 12 • |
| 13 | • ND: Lecture NDC5 - Decision models: Competitive Dynamics • PNIT: Chapter 8 - Encoding Visual Contrast |
• Question Set 13 • |
| 14 | • ND: Lecture NDC6 - Synaptic plasticity and learning • TBD |
• Question Set 14 • |
Computational neuroscience is a interdisciplinary field. To learn more about computer vision, signal processing, and deep learning, here are some beginner-friendly resources I use and recommend:
- First Principles of Computer Vision Specialization by Prof. Shree Nayar (Available on Coursera)
- The Scientist and Engineer's Guide to Digital Signal Processing by Dr. Steven W. Smith (Available Online)
- Deep Learning: A Visual Approach by Dr. Andrew Glassner (Find the Book Here)
- Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control by Prof. Steve L. Brunton and Prof. J. Nathan Kutz (Find the Book Here). Comes with excellent videos. I haven't read the second edition, though.