{"title":"Feedback Max-Margin Early Event Detector","authors":"Zhi-Fang Yang, Chin-Ting Hung","doi":"10.1109/ICASI55125.2022.9774436","DOIUrl":null,"url":null,"abstract":"The max-margin early event detector (MMED) method is the first work to solve early event detection based on the structured support vector machine(SOSVM). In this study, a feedback MMED method (FMMED) is proposed to reduce the number of constraints by introducing a feedback mechanism. The convergence of the iterative MMED method is verified to ensure the cutting plane algorithm correctly applied to the proposed FMMED method. The proposed feedback mechanism is used to find the early events related to un-tight constraints in the previous iteration, and to remove the constraints containing these early events in the current iteration. Three datasets are utilized to verify the feasibility of the proposed FMMED method.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 8th International Conference on Applied System Innovation (ICASI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICASI55125.2022.9774436","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The max-margin early event detector (MMED) method is the first work to solve early event detection based on the structured support vector machine(SOSVM). In this study, a feedback MMED method (FMMED) is proposed to reduce the number of constraints by introducing a feedback mechanism. The convergence of the iterative MMED method is verified to ensure the cutting plane algorithm correctly applied to the proposed FMMED method. The proposed feedback mechanism is used to find the early events related to un-tight constraints in the previous iteration, and to remove the constraints containing these early events in the current iteration. Three datasets are utilized to verify the feasibility of the proposed FMMED method.