RGB-D Salient Object Detection: A Survey

Revisiting Video Saliency: A Large-scale Benchmark and a New Model (CVPR18, PAMI19)

[CVPR19] DeepCO3: Deep Instance Co-segmentation by Co-peak Search and Co-saliency (Oral paper)

This Toolbox contains E-measure, S-measure, weighted F & F-measure, MAE and PR curves or bar metrics for salient object detection.

Structure-measure: A New Way to Evaluate Foreground Maps, IJCV2021 (ICCV 2017-Spotlight)

Jing Lou, Wei Zhu, Huan Wang, Mingwu Ren*, “Small Target Detection Combining Regional Stability and Saliency in a Color Image,” Multimedia Tools and Applications, vol. 76, no. 13, pp. 14781-14798, 2017.

Enhanced-alignment Measure for Binary Foreground Map Evaluation, IJCAI 2018 (Oral)

[BMVC17] Weakly Supervised Saliency Detection with A Category-Driven Map Generator

Repository for implementation of SalNet360 in Caffe

Visual saliency estimation for 360° images using stacked autoencoder.

Consistent Video Saliency using Local Gradient Flow Optimization and Global Refinement (TIP15)

The work presented explains how to segment the brain tumour area in absence of interaction with user basing his technique on a saliency map constructed from three different resonance techniques.

Finding most salient regions in images and compressing images to maintain same visual quality at lower disk space

Jing Lou, Mingwu Ren*, Huan Wang, "Regional Principal Color Based Saliency Detection," PLoS ONE, vol. 9, no. 11, pp. e112475: 1-13, 2014.

This repo shows a MATLAB implementation of score-CAM (Wang et al., CVPR workshop, 2020). This method is the first gradient-free CAM-based visualization method for explaining CNN decision.

Saliency Aware Video Object Segmentation (CVPR15, PAMI18)

RARE2007 is a feature-engineered bottom-up salienct model only using color information (no orientation)

Jing Lou*, Huan Wang, Longtao Chen, Fenglei Xu, Qingyuan Xia, Wei Zhu, Mingwu Ren*, "Exploiting Color Name Space for Salient Object Detection," Multimedia Tools and Applications, vol. 79, no. 15, pp. 10873-10897, 2020. doi:10.1007/s11042-019-07970-x

RARE2012 is a feature-engineered bottom-up visual attention model