Convolutional networks and applications in vision

Yann LeCun, Koray Kavukcuoglu, Clément Farabet

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Intelligent tasks, such as visual perception, auditory perception, and language understanding require the construction of good internal representations of the world (or "features"), which must be invariant to irrelevant variations of the input while, preserving relevant information. A major question for Machine Learning is how to learn such good features automatically. Convolutional Networks (ConvNets) are a biologicallyinspired trainable architecture that can learn invariant features. Each stage in a ConvNets is composed of a filter bank, some non-linearities, and feature pooling layers. With multiple stages, a ConvNet can learn multi-level hierarchies of features. While ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, they require large amounts of labeled training samples. We describe new unsupervised learning algorithms, and new non-linear stages that allow ConvNets to be trained with very few labeled samples. Applications to visual object recognition and vision navigation for off-road mobile robots are described.

Original languageEnglish (US)
Title of host publicationISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems
Pages253-256
Number of pages4
DOIs
StatePublished - 2010
Event2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems, ISCAS 2010 - Paris, France
Duration: May 30 2010Jun 2 2010

Other

Other2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems, ISCAS 2010
CountryFrance
CityParis
Period5/30/106/2/10

Fingerprint

Optical character recognition
Unsupervised learning
Filter banks
Object recognition
Mobile robots
Learning algorithms
Learning systems
Navigation

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

LeCun, Y., Kavukcuoglu, K., & Farabet, C. (2010). Convolutional networks and applications in vision. In ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems (pp. 253-256). [5537907] https://doi.org/10.1109/ISCAS.2010.5537907

Convolutional networks and applications in vision. / LeCun, Yann; Kavukcuoglu, Koray; Farabet, Clément.

ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems. 2010. p. 253-256 5537907.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

LeCun, Y, Kavukcuoglu, K & Farabet, C 2010, Convolutional networks and applications in vision. in ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems., 5537907, pp. 253-256, 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems, ISCAS 2010, Paris, France, 5/30/10. https://doi.org/10.1109/ISCAS.2010.5537907
LeCun Y, Kavukcuoglu K, Farabet C. Convolutional networks and applications in vision. In ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems. 2010. p. 253-256. 5537907 https://doi.org/10.1109/ISCAS.2010.5537907
LeCun, Yann ; Kavukcuoglu, Koray ; Farabet, Clément. / Convolutional networks and applications in vision. ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems. 2010. pp. 253-256
@inproceedings{4a6fcfbca1a948488297d21dd3a0835e,
title = "Convolutional networks and applications in vision",
abstract = "Intelligent tasks, such as visual perception, auditory perception, and language understanding require the construction of good internal representations of the world (or {"}features{"}), which must be invariant to irrelevant variations of the input while, preserving relevant information. A major question for Machine Learning is how to learn such good features automatically. Convolutional Networks (ConvNets) are a biologicallyinspired trainable architecture that can learn invariant features. Each stage in a ConvNets is composed of a filter bank, some non-linearities, and feature pooling layers. With multiple stages, a ConvNet can learn multi-level hierarchies of features. While ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, they require large amounts of labeled training samples. We describe new unsupervised learning algorithms, and new non-linear stages that allow ConvNets to be trained with very few labeled samples. Applications to visual object recognition and vision navigation for off-road mobile robots are described.",
author = "Yann LeCun and Koray Kavukcuoglu and Cl{\'e}ment Farabet",
year = "2010",
doi = "10.1109/ISCAS.2010.5537907",
language = "English (US)",
isbn = "9781424453085",
pages = "253--256",
booktitle = "ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems",

}

TY - GEN

T1 - Convolutional networks and applications in vision

AU - LeCun, Yann

AU - Kavukcuoglu, Koray

AU - Farabet, Clément

PY - 2010

Y1 - 2010

N2 - Intelligent tasks, such as visual perception, auditory perception, and language understanding require the construction of good internal representations of the world (or "features"), which must be invariant to irrelevant variations of the input while, preserving relevant information. A major question for Machine Learning is how to learn such good features automatically. Convolutional Networks (ConvNets) are a biologicallyinspired trainable architecture that can learn invariant features. Each stage in a ConvNets is composed of a filter bank, some non-linearities, and feature pooling layers. With multiple stages, a ConvNet can learn multi-level hierarchies of features. While ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, they require large amounts of labeled training samples. We describe new unsupervised learning algorithms, and new non-linear stages that allow ConvNets to be trained with very few labeled samples. Applications to visual object recognition and vision navigation for off-road mobile robots are described.

AB - Intelligent tasks, such as visual perception, auditory perception, and language understanding require the construction of good internal representations of the world (or "features"), which must be invariant to irrelevant variations of the input while, preserving relevant information. A major question for Machine Learning is how to learn such good features automatically. Convolutional Networks (ConvNets) are a biologicallyinspired trainable architecture that can learn invariant features. Each stage in a ConvNets is composed of a filter bank, some non-linearities, and feature pooling layers. With multiple stages, a ConvNet can learn multi-level hierarchies of features. While ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, they require large amounts of labeled training samples. We describe new unsupervised learning algorithms, and new non-linear stages that allow ConvNets to be trained with very few labeled samples. Applications to visual object recognition and vision navigation for off-road mobile robots are described.

UR - http://www.scopus.com/inward/record.url?scp=77955998889&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77955998889&partnerID=8YFLogxK

U2 - 10.1109/ISCAS.2010.5537907

DO - 10.1109/ISCAS.2010.5537907

M3 - Conference contribution

SN - 9781424453085

SP - 253

EP - 256

BT - ISCAS 2010 - 2010 IEEE International Symposium on Circuits and Systems: Nano-Bio Circuit Fabrics and Systems

ER -