Salient Objects in Clutter: Bringing Salient Object Detection to the Foreground



In this paper, we provide a comprehensive evaluation of salient object detection (SOD) models. Our analysis identifies a serious design bias of existing SOD datasets which assumes that each image contains at least one clearly outstanding salient object in low clutter. This is an unrealistic assumption. The design bias has led to a saturated high performance for state-of-the-art SOD models when evaluated on existing datasets. The models, however, still perform far from being satisfactory when applied to real-world daily scenes. Based on our analyses, we first identify 7 crucial aspects that a comprehensive and balanced dataset should fulfill. Then, we propose a new high-quality dataset and update the previous saliency benchmark. Specifically, our dataset called SOC, Salient Objects in Clutter, includes images with salient and non-salient objects from daily object categories. Beyond object category annotations, each salient image is accompanied by attributes (e.g., appearance change, clutter) that reflect common challenges in real-world scenes, and can help 1) gain a deeper insight into the SOD problem, 2) investigate the pros and cons of the SOD models, and 3) objectively assess models from different perspectives. Finally, we report attribute-based performance assessment on our SOC dataset. We believe that our dataset and results will open new directions for future research on salient object detection.


Deng-Ping Fan, Ming-Ming Cheng,  Jiang-Jiang Liu, Shang-Hua Gao,  Qibin Hou, Ali Borji

Salient Objects in Clutter: Bringing Salient Object Detection to the Foreground,  ECCV, 2018   

[project page | bib | latex | official version | 中文版pdf | 媒体报道evaluate code][ECCV poster | ECCV slides | Dataset  (730.2MB)  Baidu] [Google ]



Note that the Test Set only contains images and without ground truth.  We will create the  SOC Benchmark website soon and you can upload your result to obtain the final score in our website. Also, you can use the Validation Set as Test Set first.


If you have any question, drop us an e-mail at <>.

SOC object level detection leaderboard

[table “26” not found /]

SOC instance level detection leaderboard

…coming soon…

SOC attribute based leaderboard

…coming soon…

Traditional Methods

  1. DRFI: xxx
  2. Generic Promotion of Diffusion-Based Salient Object Detection, ICCV,2015
  3. coming soom.

CNN based Methods from 2015-current

  1. LEGS: Deep networks for saliency detection via local estimation and global search, Wang, L. et al, CVPR, 2015.
  2. MC: Saliency detection by multi-context deep learning, Zhao, R., et al, CVPR, 2015.
  3. MDF: Visual saliency based on multiscale deep features, Li, G., et al, CVPR, 2015.
  4. DCL: Deep Contrast Learning for Salient Object Detection, Li, G.,et al, CVPR, 2016.
  5. ELD: Deep Saliency with Encoded Low level Distance Map and High Level Features, Gayoung, L., et al, CVPR, 2016.
  6. DHS: DHSNet: Deep Hierarchical Saliency Network for Salient Object Detection, Liu, N., et al, CVPR, 2016.
  7. RFCN: Saliency detection with recurrent fully convolutional networks, Wang, L., et al,  ECCV, 2016.
  8. DISC: DISC: Deep image saliency computing via progressive representation learning,  Chen, T., et al,  TNNLS, 2016
  9. DS: DeepSaliency: Multi-task deep neural network model for salient object detection, Li, X., et al, TIP, 2016.
  10. IMC: Deep Salient Object Detection by Integrating Multi-level Cues, Zhang, J., et al, WACV 2017.
  11. DSS: Deeply supervised salient object detection with short connections, Hou, Q., et al, CVPR, 2017/TPAMI, 2018.
  12. NLDF: Non-local deep features for salient object detection, Luo, Z., et al, CVPR, 2017.
  13. AMU: Amulet: Aggregating multi-level convolutional features for salient object detection, Zhang, P., et al, ICCV, 2017.
  14. UCF: Learning Uncertain Convolutional Features for Accurate Saliency Detection, Zhang, P., et al, ICCV, 2017.
  15. MSR: Instance-Level Salient Object Segmentation, Li, G., et al, CVPR, 2017.
  16. MDC: 300-FPS Salient Object Detection via Minimum Directional Contrast, Huang, X., et al, TIP, 2017. [code]
  17. SRM: A stagewise refinement model for detecting salient objects in images, CVPR, 2017.
  18. Salient Object Detection using a Context-Aware Refinement Network, BMVC,2017.
  19. R3Net: Recurrent Residual Refinement Network for Saliency Detection, IJCAI,  2018.
  20. PiCANet: Learning pixel-wise contextual attention for saliency detection, CVPR, 2018.
  21. BDMPM: A bidirectional message passing model for salient object detection, CVPR, 2018.
  22. PAGRN: Progressive attention guided recurrent network for salient object detection, CVPR, 2018.
  23. DGRL: Detect globally, refine locally: A novel approach to saliency detection, CVPR, 2018.
  24. RA: Reverse Attention for Salient Object Detection, ECCV, 2018.
  25. RADF: Recurrently aggregating deep features for salient object detection, AAAI, 2018.
  26. SD: Super Diffusion for Salient Object Detection, arXive, 2018.
  27. DEF: Deep Embedding Features for Salient Object Detection, AAAI, 2019.
  28. LFRWS: Salient Object Detection with Lossless Feature Reflection and Weighted Structural Loss, TIP, 2019.
  29. SIBA: Selectivity or Invariance Boundary-aware Salient Object Detection, arXive, 2019.
  30. DRMC: Deep Reasoning with Multi-scale Context for Salient Object Detection, arXive, 2019.
  31. RDSNet: Richer and Deeper Supervision Network for Salient Object Detection, arXive, 2019.
  32. continue update….

Instance-level Salient Object Detection datasets

  1. SOC_6K (3K non-salient + 3K salient with object-level and instance-level).
  2. ILSO_1K(salient with only object-level).

Object-level Salient Object Detection datasets

Note that: If you downloaded the dataset, please cite the related citation in your paper. These datasets only for academic convenience.

Please cite related paper if you use our datasets and results

title={Salient Objects in Clutter: Bringing Salient Object Detection to the Foreground},
author={Fan, Deng-Ping and Cheng, Ming-Ming and Liu, Jiang-Jiang and Gao, Shang-Hua and Hou, Qibin and Borji, Ali},
booktitle = {European Conference on Computer Vision (ECCV)},

title={{Structure-measure: A New Way to Evaluate Foreground Maps}},
author={Fan, Deng-Ping and Cheng, Ming-Ming and Liu, Yun and Li, Tao and Borji, Ali},
booktitle = {IEEE International Conference on Computer Vision (ICCV)},
pages = {4548-4557},

author={Deng-Ping Fan, Cheng Gong, Yang Cao, Bo Ren, Ming-Ming Cheng, Ali Borji},
title={{Enhanced-alignment Measure for Binary Foreground Map Evaluation}},
booktitle={International Joint Conference on Artificial Intelligence (IJCAI)},
pages = {698-704},

7278total visits,97visits today

Leave a Reply

Your email address will not be published. Required fields are marked *