Saliency Based on Information Maximization

By Neil D. B. Bruce, John K. Tsotsos. 2005

1,278 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

On the Binding Problem in Artificial Neural Networks (2020, background): “Salient regions can be learned based on bottom-up information, such as the self-information of local image patches (Bruce and Tsotsos, 2006).”
A Deep Network Solution for Attention and Aesthetics Aware Photo Cropping (2019, methodology): “Early visual attention models [17], [18] in the vision com-”
Minimum Barrier Salient Object Detection at 80 FPS (2015, background): “Previous works in saliency detection indicate that saliency emerges from several closely related concepts such as rarity, uniqueness and local/global contrast [17, 14, 4, 2].”
Learning the Relative Importance of Objects from Tagged Images for Retrieval and Cross-Modal Search (2012, methodology): “…retrieval · Image search · Object recognition · Auto annotation · Kernelized canonical correlation analysis S.J. Hwang ( ) · K. Grauman Department of Computer Science, University of Texas at Austin, Austin, TX 78712, USA e-mail: sjhwang@cs.utexas.edu K. Grauman e-mail: grauman@cs.utexas.edu”
Deep Visual Attention Prediction (2017, methodology): “We evaluate the contribution of each component of the proposed deep attention model on TORONTO [25] and PASCAL-S [26] datasets and measure the performance using the shufﬂed AUC metric.”
RGBD Salient Object Detection via Deep Fusion (2016, methodology): “ssiﬁed into two categories: local and global approaches. Local method calculates color, edge orientation or texture contrast of a pixel/region with respect to a local window to measure saliency [14], [27]. In [14], they develop an early local based visual saliency detection method by computing cente”
Salient Object Detection in the Deep Learning Era: An In-Depth Survey (2019, methodology): “The images in the ﬁxation subset of XPIE [119] are collected from dataset in [162] and [163].”
Visual saliency estimation by nonlinearly integrating features using region covariances. (2013, methodology): “The first data set was from Bruce and Tsotsos (2006), referred to as the Toronto data set, contains 120 natural color images of size 681 · 511, each depicting an outdoor or an indoor urban scene. The eye movement data were collected from 20 subjects who free-viewed each image for 4 s. The participan”
Image Signature: Highlighting Sparse Salient Regions (2012, methodology): “To validate the saliency maps generated by our algorithm, we use the data set of human eye-tracking data introduced by Bruce and Tsotsos [1] to compare the various saliency map algorithms IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 34, NO. 1, JANUARY 2012 197 Fig.”
A Novel Multiresolution Spatiotemporal Saliency Detection Model and Its Applications in Image and Video Compression (2010, methodology): “In addition, our PQFT model can work in real time (less than 1 ms per image/frame in C/C++ implementation); however, other models ([7], [8], [14], [16]–[18]) cannot do well.”
Top-down visual saliency via joint CRF and dictionary learning (2012, background): “In the last two decades, some basic principles, such as center-surround contrast [10], self-information [3], topological connectivity [8] and spectral residual [9], have been established for computing bottom-up saliency maps, which are shown to be effective for predicting human eye movements [3, 8] ”
Learning to Detect A Salient Object (2007, methodology): “The saliency is computed by a center-surround operation [10], self-information [3], or graph-based random walk [6] using multiple features.”
Static and space-time visual saliency detection by self-resemblance. (2009, methodology): “INTRODUCTION Visual saliency detection has been of great research intere st [5], [8], [10], [14], [17], [36], [43], [42], [18], [24], [13] in recent years.”
A Comparative Study of Sampling Analysis in the Scene Classification of Optical High-Spatial Resolution Remote Sensing Imagery (2015, methodology): “- Attention by information maximization (AIM) [41]: This model is rooted in information theory, where the saliency is determined by quantifying Shannon’s self-information of each local image patch.”
Region-Based Saliency Detection and Its Application in Object Recognition (2014, methodology): “alleviate this problem, most of methods [11], [40] usually”
State-of-the-Art in Visual Attention Modeling (2013, background): “For instance, it has been claimed that visual attention is attracted to the most informative [144], the most surprising scene regions [145], or those regions that maximize reward regarding a task [109].”
Measuring the Objectness of Image Windows (2011, background): “sometimes it is unique within the image and stands out as salient [7], [21], [27], [28].”
Contextual Hypergraph Modelling for Salient Object Detection (2013, background): “Early approaches in this area focus mainly on predicting where humans look, and thus work only on human eye fixation data [1–3].”