Towards an active foveated approach to computer vision

In this paper, a series of experimental methods are presented explaining a new approach towards active foveated Computer Vision (CV). This is a collaborative effort between researchers at CONICET Mendoza Technological Scientific Center from Argentina, Argonne National Laboratory (ANL), and Loyola Un...

Descripción completa

Detalles Bibliográficos
Autores: Dematties, Dario Jesus, Rizzi, Silvio, Thiruvathukal, George, Wainselboim, Alejandro Javier
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2022
País:Argentina
Institución:Consejo Nacional de Investigaciones Científicas y Técnicas
Repositorio:CONICET Digital (CONICET)
Idioma:inglés
OAI Identifier:oai:ri.conicet.gov.ar:11336/203696
Acceso en línea:http://hdl.handle.net/11336/203696
Access Level:acceso abierto
Palabra clave:FOVEATED COMPUTER VISION
GENERAL-PURPOSE GRAPHICS PROCESSING UNITS (GPGPUS)
REINFORCEMENT LEARNING
SACCADIC BEHAVIOR
SELF-SUPERVISED LEARNING
https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
Descripción
Sumario:In this paper, a series of experimental methods are presented explaining a new approach towards active foveated Computer Vision (CV). This is a collaborative effort between researchers at CONICET Mendoza Technological Scientific Center from Argentina, Argonne National Laboratory (ANL), and Loyola University Chicago from the US. The aim is to advance new CV approaches more in line with those found in biological agents in order to bring novel solutions to the main problems faced by current CV applications. Basically this work enhances Self-supervised (SS) learning, incorporating foveated vision plus saccadic behavior in order to improve training and computational efficiency without reducing performance significantly. This paper includes a compendium of methods’ explanations, and since this is a work that is currently in progress, only preliminary results are provided. We also make our code fully available.