Abstract:High-level semantic knowledge in addition to low-level visual cues is essentially crucial for co-saliency detection. This article proposes a novel end-to-end deep learning approach for robust co-saliency detection by simultaneously learning high-level groupwise semantic representation as well as deep visual features of a given image group. The interimage interaction at the semantic level and the complementarity between the group semantics and visual features are exploited to boost the inferring capability of co-salient regions. Specifically, the proposed approach consists of a co-category learning branch and a co-saliency detection branch. While the former is proposed to learn a groupwise semantic vector using co-category association of an image group as supervision, the latter is to infer precise co-salient maps based on the ensemble of group-semantic knowledge and deep visual cues. The group-semantic vector is used to augment visual features at multiple scales and acts as a top-down semantic guidance for boosting the bottom-up inference of co-saliency. Moreover, we develop a pyramidal attention (PA) module that endows the network with the capability of concentrating on important image patches and suppressing distractions. The co-category learning and co-saliency detection branches are jointly optimized in a multitask learning manner, further improving the robustness of the approach. We construct a new large-scale co-saliency data set COCO-SEG to facilitate research of the co-saliency detection. Extensive experimental results on COCO-SEG and a widely used benchmark Cosal2015 have demonstrated the superiority of the proposed approach compared with state-of-the-art methods.

Co-saliency Detection via Sparse Reconstruction and Co-salient Object Discovery.

Co-saliency Detection Based on Hierarchical Consistency.

Co-saliency Detection Via Looking Deep and Wide

Detection of Co-salient Objects by Looking Deep and Wide

A New Method and Benchmark for Detecting Co-Saliency Within a Single Image

Co-Saliency Detection Within a Single Image

Self-supervised image co-saliency detection

Co-Saliency Detection Via Local Prediction and Global Refinement.

Re-thinking The Relations in Co-saliency Detection

A Feature-Adaptive Semi-Supervised Framework for Co-saliency Detection.

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

Co-Saliency Detection via a General Optimization Model and Adaptive Graph Learning

An Object-Oriented Visual Saliency Detection Framework Based on Sparse Coding Representations

Cosaliency Detection Based on Intrasaliency Prior Transfer and Deep Intersaliency Mining

Revisiting Co-Saliency Detection: A Novel Approach Based on Two-Stage Multi-View Spectral Rotation Co-clustering

An End-to-End Network for Co-Saliency Detection in One Single Image

Co-salient Object Detection Based on Deep Saliency Networks and Seed Propagation over an Integrated Graph

A New Algorithm for Detecting Co-Saliency in Multiple Images through Sparse Coding Representation

Salient Object Detection Employing Robust Sparse Representation and Local Consistency

Robust Deep Co-Saliency Detection With Group Semantic and Pyramid Attention

A Unified Metric Learning-Based Framework for Co-Saliency Detection.