TY - GEN
T1 - Multiscale discriminant saliency for visual attention
AU - Le Ngo, Anh Cat
AU - Ang, Kenneth Li Minn
AU - Qiu, Guoping
AU - Kah-Phooi, Jasmine Seng
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2013.
PY - 2013
Y1 - 2013
N2 - The bottom-up saliency, an early stage of humans' visual attention, can be considered as a binary classification problem between center and surround classes. Discriminant power of features for the classification is measured as mutual information between features and two classes distribution. The estimated discrepancy of two feature classes very much depends on considered scale levels; then, multi-scale structure and discriminant power are integrated by employing discrete wavelet features and Hidden markov tree (HMT). With wavelet coefficients and Hidden Markov Tree parameters, quad-tree like label structures are constructed and utilized in maximum a posterior probability (MAP) of hidden class variables at corresponding dyadic sub-squares. Then, saliency value for each dyadic square at each scale level is computed with discriminant power principle and the MAP. Finally, across multiple scales is integrated the final saliency map by an information maximization rule. Both standard quantitative tools such as NSS, LCC, AUC and qualitative assessments are used for evaluating the proposed multiscale discriminant saliency method (MDIS) against the well-know information-based saliency method AIM on its Bruce Database wity eye-tracking data. Simulation results are presented and analyzed to verify the validity of MDIS as well as point out its disadvantages for further research direction.
AB - The bottom-up saliency, an early stage of humans' visual attention, can be considered as a binary classification problem between center and surround classes. Discriminant power of features for the classification is measured as mutual information between features and two classes distribution. The estimated discrepancy of two feature classes very much depends on considered scale levels; then, multi-scale structure and discriminant power are integrated by employing discrete wavelet features and Hidden markov tree (HMT). With wavelet coefficients and Hidden Markov Tree parameters, quad-tree like label structures are constructed and utilized in maximum a posterior probability (MAP) of hidden class variables at corresponding dyadic sub-squares. Then, saliency value for each dyadic square at each scale level is computed with discriminant power principle and the MAP. Finally, across multiple scales is integrated the final saliency map by an information maximization rule. Both standard quantitative tools such as NSS, LCC, AUC and qualitative assessments are used for evaluating the proposed multiscale discriminant saliency method (MDIS) against the well-know information-based saliency method AIM on its Bruce Database wity eye-tracking data. Simulation results are presented and analyzed to verify the validity of MDIS as well as point out its disadvantages for further research direction.
UR - http://www.scopus.com/inward/record.url?scp=85099486055&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-39637-3_37
DO - 10.1007/978-3-642-39637-3_37
M3 - Conference Proceeding
AN - SCOPUS:85099486055
SN - 9783642396366
VL - 7971
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 464
EP - 484
BT - Computational Science and Its Applications, ICCSA 2013 - 13th International Conference, Proceedings
PB - Springer Verlag
T2 - 13th International Conference on Computational Science and Its Applications, ICCSA 2013
Y2 - 24 June 2013 through 27 June 2013
ER -