Large-Scale Optimization of Hierarchical Features for Saliency Prediction in Natural Images
- 1 June 2014
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2798-2805
- https://doi.org/10.1109/cvpr.2014.358
Abstract
Saliency prediction typically relies on hand-crafted (multiscale) features that are combined in different ways to form a "master" saliency map, which encodes local image conspicuity. Recent improvements to the state of the art on standard benchmarks such as MIT1003 have been achieved mostly by incrementally adding more and more hand-tuned features (such as car or face detectors) to existing models. In contrast, we here follow an entirely automatic data-driven approach that performs a large-scale search for optimal features. We identify those instances of a richly-parameterized bio-inspired model family (hierarchical neuromorphic networks) that successfully predict image saliency. Because of the high dimensionality of this parameter space, we use automated hyperparameter optimization to efficiently guide the search. The optimal blend of such multilayer features combined with a simple linear classifier achieves excellent performance on several image saliency benchmarks. Our models outperform the state of the art on MIT1003, on which features and classifiers are learned. Without additional training, these models generalize well to two other image saliency data sets, Toronto and NUSEF, despite their different image content. Finally, our algorithm scores best of all the 23 models evaluated to date on the MIT300 saliency challenge, which uses a hidden test set to facilitate an unbiased comparison.Keywords
This publication has 16 references indexed in Scilit:
- Learning a saliency map using fixated locations in natural scenesJournal of Vision, 2011
- Fast and Efficient Saliency Detection Using Sparse Sampling and Kernel Density EstimationLecture Notes in Computer Science, 2011
- A High-Throughput Screening Approach to Discovering Good Forms of Biologically Inspired Visual RepresentationPLoS Computational Biology, 2009
- Learning to predict where humans lookPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2009
- SUN: A Bayesian framework for saliency using natural statisticsJournal of Vision, 2008
- Saliency Detection: A Spectral Residual ApproachPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2007
- Bottom-up saliency is a discriminant processPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2007
- Components of bottom-up gaze allocation in natural imagesVision Research, 2005
- The Boosting Approach to Machine Learning: An OverviewPublished by Springer Science and Business Media LLC ,2003
- A model of saliency-based visual attention for rapid scene analysisIEEE Transactions on Pattern Analysis and Machine Intelligence, 1998