
Texture Perception (cont'd)
MIT professor Josh McDermott continues his lecture on texture perception for course 9.35, Perception, examining how the visual and auditory systems recognize texture. The lecture builds on prior material, working through the computational and psychophysical basis for how humans identify visual textures like fabrics or surfaces and auditory textures like rain or applause. McDermott discusses models that extract statistical summary representations from sensory input, arguing that the brain relies on such summary statistics rather than tracking every individual element in a scene or sound. The seventy four minute session follows the format of an MIT lecture hall talk, with McDermott working through examples and diagrams to connect perceptual experiments to underlying computational theories. It sits within a graduate level course on perception at MIT, part of a series covering vision, hearing, and the general problem of how the brain organizes raw sensory data into recognizable categories.