{"title":"Active visual search (AViS) dataset Track: Sensory exploitation","authors":"Alexandra Hildenbrandt, Bernard Abayowa","doi":"10.1109/NAECON.2015.7443101","DOIUrl":null,"url":null,"abstract":"Image retrieval is a vital function for an active vision system recognizing vehicles in dynamic scenes. A pan-tilt-zoom camera provides a range of retrieval options, but is somewhat limited for algorithm development purposes in that if the conditions are not within a certain operative range (due to very bad weather, poor lighting conditions, etc.), retrieved images are not likely to be very useful. Therefore it becomes imperative to have an alternative source during development. This paper presents a dataset with an associated application programming interface (API) which seeks to be that alternative. In addition to simulating the actions and capabilities of pan-tilt-zoom cameras, this dataset and API provide access to template vehicle images as well as information on verified locations of targets which can be used for training and testing purposes.","PeriodicalId":133804,"journal":{"name":"2015 National Aerospace and Electronics Conference (NAECON)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2015-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 National Aerospace and Electronics Conference (NAECON)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/NAECON.2015.7443101","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
Image retrieval is a vital function for an active vision system recognizing vehicles in dynamic scenes. A pan-tilt-zoom camera provides a range of retrieval options, but is somewhat limited for algorithm development purposes in that if the conditions are not within a certain operative range (due to very bad weather, poor lighting conditions, etc.), retrieved images are not likely to be very useful. Therefore it becomes imperative to have an alternative source during development. This paper presents a dataset with an associated application programming interface (API) which seeks to be that alternative. In addition to simulating the actions and capabilities of pan-tilt-zoom cameras, this dataset and API provide access to template vehicle images as well as information on verified locations of targets which can be used for training and testing purposes.