Representation is representation of similarities Edelman S Behavioral
Representation is representation of similarities Edelman S. , ‘Behavioral and Brain Sciences, ’ 1998. Naresh P. Cuntoor
Introduction Naresh P. Cuntoor
Groundwork w Distal shape space w Proximal shape space w Representing similarity – distinctness, NN preservation, full similarity spectrum w Distal to proximal mapping, M - constraints and composition, distance rank preservation Naresh P. Cuntoor
Analysis of Mapping f 1(p): Geometry f 2(p, z): Imaging f 3(p, z): Measurements f 4(p): Dimensionality reduction • f 4 and f 3 need to counteract the z-dependency of f 2 • Absolute invariance not necessary –Need: influence of shape space changes > view space changes Naresh P. Cuntoor
Representation = Measurement + Dimensionality Reduction • Another example: Sarkar’s face space to affine space Naresh P. Cuntoor
Chorus of prototypes w An ensemble of tuned classifiers w Smooth response degradation Naresh P. Cuntoor
Similarity w Levels: basic, subordinate, superordinate w Features of similarity: pi(A): ith classifier w Measures of similarity Naresh P. Cuntoor
What’s the brain doing? w Novel objects – how chorus deals with it w New Pandemonium –feature demons, cognitive demons, master demons w ‘Democracy’ in chorus w Perception of similarity – ppl. classify maps Naresh P. Cuntoor
Experiments and Predictions w To test second order isormophism w Computer rendered 3 D animal shapes and nonsense shapes w Predicting distortion in the MDS setting – can chorus do well? w Parameter space distances – some more important? w Priming – how does it affect? w Neurobiology – columns in IT cortex w Qualia – attributes of objects w Scene richness – humans don’t see everything w Bottom-up vs. top-down – Sinha’s STICKS approach Naresh P. Cuntoor
Challenges Naresh P. Cuntoor
- Slides: 10