Abstract:Saliency detection is an important and challenging research topic due to the variety and complex of the background and saliency regions. In this paper, we present a novel unsupervised saliency detection approach by exploiting a learning-based ranking framework. First, the local linear regression model is adopted to simulate the local manifold structure of every image element, which is approximately linear. Using the background queries from the boundary prior, we construct a unified objective function to globally minimize all the errors of the local models for the whole image element points. The Laplacian matrix is learned via optimizing the unified objective function. Low-level image features as well as high-level semantic information extracted from deep neural networks are used for the Laplacian matrix learning. Based on the learnt Laplacian matrix, the saliency of the image element is measured as the relevance ranking to the background queries. The foreground queries are obtained from the background-based saliency and the relevance ranking to the foreground queries is calculated in the same way as the background-based saliency. Second, we calculate an enhanced similarity matrix by fusing two different-level deep feature metrics through cross diffusion. A propagation algorithm uses this enhanced similarity matrix to better exploit the intrinsic relevance of similar regions and improve the saliency ranking results effectively. Results on four benchmark datasets with pixel-wise accurate labelling demonstrate that the proposed unsupervised method shows better performance compared with the newest state-of-the-art methods and is competitive with deep learning-based methods.

Robust Ranking on Manifold for Salient Object Detection

SaliencyRank: Two-stage Manifold Ranking for Salient Object Detection.

Salient Object Detection Via Graph-Based Flexible Manifold Ranking

An Improved Graph-based Manifold Ranking for Salient Object Detection

Salient Object Detection Based on Boundary Contrast with Regularized Manifold Ranking

Saliency Detection Based on Robust Foreground Selection

Eigenobject-wise saliency detection based on manifold ranking.

Salient Object Detection Based on Multi-Feature Graphs and Improved Manifold Ranking

Salient Object Detection Via Incorporating Multiple Manifold Ranking

Saliency Detection Via a Multiple Self-Weighted Graph-Based Manifold Ranking.

Saliency Detection Via Manifold Ranking on Multi-Layer Graph

Saliency Detection Via Graph-Based Manifold Ranking

Saliency Detection Based on Spread Pattern and Manifold Ranking.

A New Graph Ranking Model for Image Saliency Detection Problem

Visual Saliency Detection Via Prior Regularized Manifold Ranking.

Hybrid of Extended Locality-Constrained Linear Coding and Manifold Ranking for Salient Object Detection.

Saliency Detection via Multi-view Synchronized Manifold Ranking.

Salient object detection with low-rank approximation and ℓ2,1-norm minimization

Saliency Detection: Multi-Level Combination Approach Via Graph-Based Manifold Ranking

Local Regression Ranking for Saliency Detection

Saliency Ranker: A New Salient Object Detection Method