{"id":13883,"date":"2025-04-10T13:24:15","date_gmt":"2025-04-10T18:24:15","guid":{"rendered":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/?p=13883"},"modified":"2026-04-15T08:35:56","modified_gmt":"2026-04-15T12:35:56","slug":"multi-task-learning-with-multi-annotation-triplet-loss-for-improved-object-detection","status":"publish","type":"post","link":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/2025\/04\/10\/multi-task-learning-with-multi-annotation-triplet-loss-for-improved-object-detection\/","title":{"rendered":"Multi-Task Learning with Multi-Annotation Triplet Loss for Improved Object Detection"},"content":{"rendered":"<h2>Abstract:<\/h2>\n<p>Triplet loss traditionally relies only on class labels and does not use all available information in multi-task scenarios where multiple types of annotations are available. This paper introduces a Multi-Annotation Triplet Loss (MATL) framework that extends triplet loss by incorporating additional annotations, such as bounding box information, alongside class labels in the loss formulation. By using these complementary annotations, MATL improves multi-task learning for tasks requiring both classification and localization. Experiments on an aerial wildlife imagery dataset demonstrate that MATL outperforms conventional triplet loss in both classification and localization. These findings highlight the benefit of using all available annotations for triplet loss in multi-task learning frameworks.<\/p>\n<h2>Links:<\/h2>\n<p><a href=\"https:\/\/arxiv.org\/abs\/2504.08054\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-470 size-full\" src=\"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-content\/uploads\/sites\/759\/2016\/09\/arxiv.png\" alt=\"arXiv Interactive Segmentation With Prototype Learning for Few-Shot Root Annotation\" width=\"90\" height=\"90\" \/><\/a><\/p>\n<h2>Citation:<\/h2>\n<pre><code>M. Zhou, A. Dutt, and A. Zare, \"Multi-Task Learning with Multi-Annotation Triplet Loss for Improved Object Detection,\" in IGARSS 2025 - 2025 IEEE International Geoscience and Remote Sensing Symposium, 2025.<\/code><\/pre>\n<pre class=\"verbatim select-on-click\" title=\"click to copy to clipboard\"><code>@inproceedings{zhou2025multi-task,\ntitle={Multi-Task Learning with Multi-Annotation Triplet Loss for Improved Object Detection},\nauthor={M. Zhou, A. Dutt, and A. Zare},\nbooktitle={IGARSS 2025-2025 IEEE International Geoscience and Remote Sensing Symposium},\nyear={2025},\norganization={IEEE}\n}\n<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"<p>Abstract: Triplet loss traditionally relies only on class labels and does not use all available information in multi-task scenarios where multiple types of annotations are available. This paper introduces a Multi-Annotation Triplet Loss (MATL) framework that extends triplet loss by incorporating additional annotations, such as bounding box information, alongside class labels in the loss formulation. [&hellip;]<\/p>\n","protected":false},"author":28,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"single-templates\/single-sidebar-none.php","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"featured_post":"off","footnotes":"","_links_to":"","_links_to_target":""},"categories":[17,13],"tags":[65,133,151,169,223,427,453,475,481,623,759,803],"class_list":["post-13883","post","type-post","status-publish","format-standard","hentry","category-conference_paper","category-publication","tag-aerial-imagery","tag-bounding-box-annotations","tag-classification","tag-computer-vision","tag-deep-learning","tag-localization","tag-metric-learning","tag-multi-annotation-learning","tag-multi-task-learning","tag-representation-learning","tag-triplet-loss","tag-wildlife-monitoring"],"acf":[],"_links":{"self":[{"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/posts\/13883","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/users\/28"}],"replies":[{"embeddable":true,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/comments?post=13883"}],"version-history":[{"count":2,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/posts\/13883\/revisions"}],"predecessor-version":[{"id":16579,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/posts\/13883\/revisions\/16579"}],"wp:attachment":[{"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/media?parent=13883"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/categories?post=13883"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/faculty.eng.ufl.edu\/machine-learning\/wp-json\/wp\/v2\/tags?post=13883"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}