Opening Object Recognition (cont'd), Texture Perception
Object Recognition (cont'd), Texture Perception
- Length
- 1 h 20 min
- Level
- Undergraduate
- Recorded
- MIT2026
Chapter 1 of 8: Challenges and models of object recognition
Reads object identity linearly from IT populations, then predicts texture boundaries from filter energy.
Object recognition in the ventral visual stream is the starting point for McDermott's session 18 of MIT 9.35. He shows why image variability makes recognition hard, then cites Hung, Kreiman, Poggio and DiCarlo (2005) to show how IT populations make object identity linearly separable. Face patches, representational dissimilarity matrices and encoding models built from convolutional network features serve to compare brains across species and against trained networks. The second half turns to texture as a cue to material, depth and scene segmentation. Squaring and smoothing early filter responses yields energy measures that predict texture boundary salience, and subband filter histograms characterize textures by their average statistics. The session assumes familiarity with V1, V4, IT, receptive fields, convolution and gradient descent.
Object recognition in the ventral visual stream is the starting point for McDermott's session 18 of MIT 9.35. He shows why image variability makes recognition hard, then cites Hung, Kreiman, Poggio and DiCarlo (2005) to show how IT populations make object identity linearly separable. Face patches, representational dissimilarity matrices and encoding models built from convolutional network features serve to compare brains across species and against trained networks. The second half turns to texture as a cue to material, depth and scene segmentation. Squaring and smoothing early filter responses yields energy measures that predict texture boundary salience, and subband filter histograms characterize textures by their average statistics. The session assumes familiarity with V1, V4, IT, receptive fields, convolution and gradient descent.
Farah (1995)
Thorpe et al. (1996) Nature
Kobatake et al. (1994)
Logothetis et al. (1995)
DiCarlo and Cox (2007)
Hung, Kreiman, Poggio, and DiCarlo (2005) Science
This lecture
Lecture 1 of 6 of Sensation and Perception in Cognitive Neuroscience.
Choose a node to see how it connects to this lecture. What comes before is on the leftat the top. What comes after is on the rightat the bottom. The ring is what it teaches. People and places are above.
Press a chosen node again to open it.
Your comment is kept while you sign up.Sign in
Rapid, explicit object representation in IT
A map of face patches in IT
A neural map of face selectivity across IT
Brain Imaging on face recognition
Case study of Dr. P
Pallis, 1955
Kriegeskorte et al. 2008
Yamins et al. 2014
Eickenberg et al. 2017
Kell et al. 2018
Early vision and texture perception
Heeger and Bergen 1995
Freeman & Adelson, 91