CSAIL Event Calendar: Previous Series

Abstract:
The human visual system contains 10^11 neurons, of which one third are devoted to vision alone. We know the ‘wiring diagram’ and the psychophysics of the system quite well, but we don’t understand the meaning of the inter-neuronal signals. Since each neuron receives trains of random voltage spikes from about 10^4 other neurons, the traffic is noisy and intense and yet we see the world as stable, clear, and gapless. I will summarize the wiring diagram (anatomy), the conventional ‘bag of tricks’ theory of how the system works (which I think is wrong), and will outline our alternative model, the Excitable Neuronal Array, which is a simple neural network characterized by a small number of parameters. The dynamic properties of the ENA can account for perception of motion, depth, and perhaps other visual abilities. Finally I will show some static pictures in which illusory motion is seen. Our model accounts for these illusions as a consequence of statistical voltage fluctuations, analogous to stochastic resonance in a noisy brain, in which noise is otherwise nearly always suppressed and hidden from conscious perception.

Biography:
Donald A. Glaser studied high energy physics and invented the bubble chamber for visualizing and studying the properties of elementary particles of physics. He received the Nobel Prize for this work in 1960 at the age of 34. He later worked in molecular biology and microbial genetics, inventing a number of new methods for automating large scale hunts for valuable mutants and other experimental tasks. He co-founded the very first biotech company – Chiron. Besides the Nobel Prize, Dr. Glaser’s awards include the Elliot Cresson Medal, the Gold Medal Award from the Case Institute of Technology, a Distinguished Research Fellow designation from the Smith-Kettlewell Institute for Vision Research, and the Golden Plate Award from the American Academy of Achievement.

Glaser’s current research has shifted to the construction of computational models that shed light on the physics and physiology of human perception. His goal is to construct computational models of the human visual system to explain its performance in terms of its physiology and anatomy. Simulations are essential since conventional mathematics is ill-suited to building a useful bridge between psychophysics and neurobiology.