A novel hierarchical model of attention: maximizing information acquisition

Authors:
Yang Cao;Liqing Zhang
Affiliations:
MOE-Microsoft Laboratory for Intelligent Computing and Intelligent Systems, and Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China;MOE-Microsoft Laboratory for Intelligent Computing and Intelligent Systems, and Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China
Venue:
ACCV'09 Proceedings of the 9th Asian conference on Computer Vision - Volume Part I
Year:
2009

Citing 5
Cited 1

A Model of Saliency-Based Visual Attention for Rapid Scene Analysis

IEEE Transactions on Pattern Analysis and Machine Intelligence
Saliency, Scale and Image Description

International Journal of Computer Vision
Metabolically Efficient Information Processing

Neural Computation
Detecting Irregularities in Images and in Video

International Journal of Computer Vision
Decision-theoretic saliency: Computational principles, biological plausibility, and implications for neurophysiology and psychophysics

Neural Computation

Attention Focusing Model for Nexting Based on Learning and Reasoning

Proceedings of the 2010 conference on Biologically Inspired Cognitive Architectures 2010: Proceedings of the First Annual Meeting of the BICA Society

Quantified Score

Hi-index	0.00

Visualization

Abstract

A visual attention system should preferentially locate the most informative spots in complex environments. In this paper, we propose a novel attention model to produce saliency maps by generating information distributions on incoming images. Our model automatically marks spots with large information amount as saliency, which ensures the system gains the maximum information acquisition through attending these spots. By building a biological computational framework, we use the neural coding length as the estimation of information, and introduce relative entropy to simplify this calculation. Additionally, a real attention system should be robust to scales. Inspired by the visual perception process, we design a hierarchical framework to handle multi-scale saliency. From experiments we demonstrated that the proposed attention model is efficient and adaptive. In comparison to mainstream approaches, our model achieves better accuracy on fitting human fixations.