Eccentricity dependent deep neural networks: Modeling invariance in human vision

Humans can recognize objects in a way that is invariant to scale, translation, and clutter. We use invariance theory as a conceptual basis, to computationally model this phenomenon. This theory discusses the role of eccentricity in human visual processing, and is a generalization of feedforward conv...

Full description

Bibliographic Details
Main Authors: Chen, Francis X. (Contributor), Roig Noguera, Gemma (Contributor), Isik, Leyla (Contributor), Boix Bosch, Xavier (Contributor), Poggio, Tomaso A (Contributor)
Other Authors: Center for Brains, Minds, and Machines (Contributor), Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science (Contributor)
Format: Article
Language:English
Published: Association for the Advancement of Artificial Intelligence, 2017-11-22T16:03:27Z.
Subjects:
Online Access:Get fulltext
LEADER 01803 am a22002653u 4500
001 112279
042 |a dc 
100 1 0 |a Chen, Francis X.  |e author 
100 1 0 |a Center for Brains, Minds, and Machines  |e contributor 
100 1 0 |a Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science  |e contributor 
100 1 0 |a Chen, Francis X.  |e contributor 
100 1 0 |a Roig Noguera, Gemma  |e contributor 
100 1 0 |a Isik, Leyla  |e contributor 
100 1 0 |a Boix Bosch, Xavier  |e contributor 
100 1 0 |a Poggio, Tomaso A  |e contributor 
700 1 0 |a Roig Noguera, Gemma  |e author 
700 1 0 |a Isik, Leyla  |e author 
700 1 0 |a Boix Bosch, Xavier  |e author 
700 1 0 |a Poggio, Tomaso A  |e author 
245 0 0 |a Eccentricity dependent deep neural networks: Modeling invariance in human vision 
260 |b Association for the Advancement of Artificial Intelligence,   |c 2017-11-22T16:03:27Z. 
856 |z Get fulltext  |u http://hdl.handle.net/1721.1/112279 
520 |a Humans can recognize objects in a way that is invariant to scale, translation, and clutter. We use invariance theory as a conceptual basis, to computationally model this phenomenon. This theory discusses the role of eccentricity in human visual processing, and is a generalization of feedforward convolutional neural networks (CNNs). Our model explains some key psychophysical observations relating to invariant perception, while maintaining important similarities with biological neural architectures. To our knowledge, this work is the first to unify explanations of all three types of invariance, all while leveraging the power and neurological grounding of CNNs. 
655 7 |a Article 
773 |t 2017 AAAI Spring Symposium Series, Science of Intelligence: Computational Principles of Natural and Artificial Intelligence