{"662806":{"#nid":"662806","#data":{"type":"event","title":"Phd Defense by Sungtae An","body":[{"value":"\u003Cp\u003E\u003Cstrong\u003ETitle\u003C\/strong\u003E: Weakly Supervised Deep Learning for Human Activity Recognition\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003ESungtae An\u003C\/strong\u003E\u003C\/p\u003E\r\n\r\n\u003Cp\u003EPh.D. Candidate in Computer Science\u003C\/p\u003E\r\n\r\n\u003Cp\u003ESchool of Interactive Computing\u003C\/p\u003E\r\n\r\n\u003Cp\u003EGeorgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003EDate\u003C\/strong\u003E: Wednesday, November 16, 2022\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003ETime\u003C\/strong\u003E: 3:00 PM - 5:00 PM EST\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003ELocation\u003C\/strong\u003E: TSRB 523A\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003ECommittee\u003C\/strong\u003E:\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. Omer T. Inan (Advisor), School of Electrical and Computer Engineering, Georgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. James Rehg (Co-advisor), School of Interactive Computing, Georgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. Thomas Ploetz, School of Interactive Computing, Georgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. Jon Duke, School of Interactive Computing, Georgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. Mindy L. Millard-Stafford, School of Biological Sciences, Georgia Institute of Technology\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDr. Alessio Medda, Aerospace, Transportation \u0026amp; Advanced Systems Laboratory, Georgia Tech Research Institute\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u003Cstrong\u003EAbstract\u003C\/strong\u003E:\u003C\/p\u003E\r\n\r\n\u003Cp\u003EHuman activity recognition (HAR) using wearable sensors and machine learning algorithms is an emerging capability in domains including but not limited to healthcare and ergometric analysis of populations by providing context to physiological measures from wearable sensors during natural daily living activities.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDespite the success of deep supervised models in recent years, obtaining a fully labeled HAR dataset is often challenging due to the high cost and workforce associated with labeling.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EIn this dissertation, I address the challenge with the following contributions.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EFirst, I present the bilateral domain adaptation problem in HAR for the first time and propose AdaptNet, a semi-supervised deep translation network.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EAdaptNet enables information fusion of two different data domains using both unlabeled and labeled data.\u003C\/p\u003E\r\n\r\n\u003Cp\u003ENext, I present a novel framework, DynaLAP, a semi-supervised variational recurrent neural network with a dynamic prior distribution, to perform activity recognition in fixed routes and protocols.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EDynaLAP implicitly exploits the information about the environment to enhance HAR in fixed protocols such as military and athletic training with few labeled subject data.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EThen, I present DualCPC, self-supervised pre-training with the contrastive predictive coding framework using a tri-axial accelerometer signal and corresponding physiological variable measurements such as instantaneous oxygen uptake (VO2) during activities performed.\u003C\/p\u003E\r\n\r\n\u003Cp\u003EUtilizing the additional physiological variable in training time only, the DualCPC pre-trained model outperformed the baseline models across different numbers of labeled training data available.\u0026nbsp;\u003C\/p\u003E\r\n\r\n\u003Cp\u003EFinally, I conclude the dissertation by discussing the potential future work and extensions.\u003C\/p\u003E\r\n\r\n\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\r\n","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":"","field_summary_sentence":[{"value":"Weakly Supervised Deep Learning for Human Activity Recognition"}],"uid":"27707","created_gmt":"2022-11-03 11:32:22","changed_gmt":"2022-11-03 11:32:22","author":"Tatianna Richardson","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2022-11-16T15:00:00-05:00","event_time_end":"2022-11-16T17:00:00-05:00","event_time_end_last":"2022-11-16T17:00:00-05:00","gmt_time_start":"2022-11-16 20:00:00","gmt_time_end":"2022-11-16 22:00:00","gmt_time_end_last":"2022-11-16 22:00:00","rrule":null,"timezone":"America\/New_York"},"extras":[],"groups":[{"id":"221981","name":"Graduate Studies"}],"categories":[],"keywords":[{"id":"100811","name":"Phd Defense"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[{"id":"1788","name":"Other\/Miscellaneous"}],"invited_audience":[{"id":"78761","name":"Faculty\/Staff"},{"id":"78771","name":"Public"},{"id":"78751","name":"Undergraduate students"}],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}