{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deep-saliency-with-encoded-low-level-distance","title":"Deep Saliency with Encoded Low level Distance Map and High Level Features","arxiv_id":"1604.05495","date":"2016-04-19","proceeding":"CVPR 2016 6","authors":["Gayoung Lee","Yu-Wing Tai","Junmo Kim"],"abstract":"Recent advances in saliency detection have utilized deep learning to obtain\nhigh level features to detect salient regions in a scene. These advances have\ndemonstrated superior results over previous works that utilize hand-crafted low\nlevel features for saliency detection. In this paper, we demonstrate that\nhand-crafted features can provide complementary information to enhance\nperformance of saliency detection that utilizes only high level features. Our\nmethod utilizes both high level and low level features for saliency detection\nunder a unified deep learning framework. The high level features are extracted\nusing the VGG-net, and the low level features are compared with other parts of\nan image to form a low level distance map. The low level distance map is then\nencoded using a convolutional neural network(CNN) with multiple 1X1\nconvolutional and ReLU layers. We concatenate the encoded low level distance\nmap and the high level features, and connect them to a fully connected neural\nnetwork classifier to evaluate the saliency of a query region. Our experiments\nshow that our method can further improve the performance of state-of-the-art\ndeep learning-based saliency detection methods.","url_abs":"http://arxiv.org/abs/1604.05495v1","url_pdf":"http://arxiv.org/pdf/1604.05495v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deep-saliency-with-encoded-low-level-distance","repo_url":"https://github.com/gylee1103/ELDNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"deep-saliency-with-encoded-low-level-distance","repo_url":"https://github.com/gylee1103/SaliencyELD","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"saliency-detection","task_name":"Saliency Detection"}],"methods":[{"method_slug":"relu","method_name":"ReLU"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1604.05495","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}