{"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/hierarchical-spatial-aware-siamese-network","title":"Hierarchical Spatial-aware Siamese Network for Thermal Infrared Object Tracking","arxiv_id":"1711.09539","date":"2017-11-27","proceeding":null,"authors":["Xin Li","Qiao Liu","Nana Fan","Zhenyu He","Hongzhi Wang"],"abstract":"Most thermal infrared (TIR) tracking methods are discriminative, treating the\ntracking problem as a classification task. However, the objective of the\nclassifier (label prediction) is not coupled to the objective of the tracker\n(location estimation). The classification task focuses on the between-class\ndifference of the arbitrary objects, while the tracking task mainly deals with\nthe within-class difference of the same objects. In this paper, we cast the TIR\ntracking problem as a similarity verification task, which is coupled well to\nthe objective of the tracking task. We propose a TIR tracker via a Hierarchical\nSpatial-aware Siamese Convolutional Neural Network (CNN), named HSSNet. To\nobtain both spatial and semantic features of the TIR object, we design a\nSiamese CNN that coalesces the multiple hierarchical convolutional layers.\nThen, we propose a spatial-aware network to enhance the discriminative ability\nof the coalesced hierarchical feature. Subsequently, we train this network end\nto end on a large visible video detection dataset to learn the similarity\nbetween paired objects before we transfer the network into the TIR domain.\nNext, this pre-trained Siamese network is used to evaluate the similarity\nbetween the target template and target candidates. Finally, we locate the\ncandidate that is most similar to the tracked target. Extensive experimental\nresults on the benchmarks VOT-TIR 2015 and VOT-TIR 2016 show that our proposed\nmethod achieves favourable performance compared to the state-of-the-art\nmethods.","url_abs":"http://arxiv.org/abs/1711.09539v2","url_pdf":"http://arxiv.org/pdf/1711.09539v2.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":"hierarchical-spatial-aware-siamese-network","repo_url":"https://github.com/QiaoLiuHit/HSSNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"object-tracking","task_name":"Object Tracking"},{"task_slug":"thermal-infrared-object-tracking","task_name":"Thermal Infrared Object Tracking"}],"methods":[{"method_slug":"siamese-network","method_name":"Siamese Network"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}