Next-Generation Image and Sound Processing Strategies: Exploiting the Biological Model

Abstract

The main objective of this project is to extend the technical state-of-the-art in mid-level visual and auditory signal processing using an integrative biologically inspired approach. Though our research and development efforts are focused on different levels of sensory information processing, from low-level sensory adaptation to object selection and recognition, all of our efforts described in this progress report intersect at the level of features: what low level sensory features to extract, what methods used to extract them, and how to adapt feature detector parameters, such as their gains, in order to perform optimally in changing environments. In this progress report, we detail developments in each of these areas.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 2007
Accession Number
ADA462740

Entities

People

  • Laurent Itti
  • Mel W. Bartlett
  • Norberto M. Grzywacz
  • Shrikanth Narayanan

Organizations

  • University of Southern California

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Auditory Signals
  • Automated Speech Recognition
  • Brain
  • Climate Change
  • Computer Vision
  • Contrast
  • Detectors
  • Dimensionality Reduction
  • Environment
  • Feature Extraction
  • Information Processing
  • Information Science
  • Kalman Filtering
  • Object Recognition
  • Recognition
  • Signal Processing
  • Visual Servoing

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Robotics and Automation.
  • Technical Research and Report Writing.