{"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/training-domain-specific-models-for-energy","title":"Training Domain Specific Models for Energy-Efficient Object Detection","arxiv_id":"1811.02689","date":"2018-11-06","proceeding":null,"authors":["Kentaro Yoshioka","Edward Lee","Mark Horowitz"],"abstract":"We propose an end-to-end framework for training domain specific models (DSMs)\nto obtain both high accuracy and computational efficiency for object detection\ntasks. DSMs are trained with distillation \\cite{hinton2015distilling} and focus\non achieving high accuracy at a limited domain (e.g. fixed view of an\nintersection). We argue that DSMs can capture essential features well even with\na small model size, enabling higher accuracy and efficiency than traditional\ntechniques. In addition, we improve the training efficiency by reducing the\ndataset size by culling easy to classify images from the training set. For the\nlimited domain, we observed that compact DSMs significantly surpass the\naccuracy of COCO trained models of the same size. By training on a compact\ndataset, we show that with an accuracy drop of only 3.6\\%, the training time\ncan be reduced by 93\\%. The codes are uploaded in\nhttps://github.com/kentaroy47/training-domain-specific-models.","url_abs":"http://arxiv.org/abs/1811.02689v2","url_pdf":"http://arxiv.org/pdf/1811.02689v2.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":"training-domain-specific-models-for-energy","repo_url":"https://github.com/kentaroy47/training-domain-specific-models","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"training-domain-specific-models-for-energy","repo_url":"https://github.com/kentaroy47/DatasetCulling","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"training-domain-specific-models-for-energy","repo_url":"https://github.com/skiteskopes/FID_CULLING","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"training-domain-specific-models-for-energy","repo_url":"https://github.com/skiteskopes/datasetCulling","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"computational-efficiency","task_name":"Computational Efficiency"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}