SuppleView

Proceedings of the International Conference on Advanced Visual Interfaces Pub Date : 2020-09-28 DOI:10.1145/3399715.3401952

Natsuki Hamanishi, Junichi Rekimoto

{"title":"SuppleView","authors":"Natsuki Hamanishi, Junichi Rekimoto","doi":"10.1145/3399715.3401952","DOIUrl":null,"url":null,"abstract":"In this paper, we proposed the rotation based browsing method for video learning in personal training. SuppleView, which is flexible in respect of the user's physical position while viewing a video, enables coordinate translation free viewing between an observer and an actor. Previous work on video learning have not enough explored the limitation on the observation angle, although its angle effects for observer's comprehension and caused only in video learning not in observation with the actual trainer. The method solve this basic limitation by inferring the 3D pose of frames in a video. Based on those poses, we create an virtual agent with 3D model as an actor of movements, that is same with the movement in an original 2D video. The system transition for the two actors depends on the physical rotation of the user's head so that the angle of view for observing the actor also changes. Hence, the content rendering in proposed viewer could be provided to trainees as in kind-full form for their observation in the point of an observation angle of view. We report the method overview and our prototyping to show the proof of concept.","PeriodicalId":149902,"journal":{"name":"Proceedings of the International Conference on Advanced Visual Interfaces","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2020-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the International Conference on Advanced Visual Interfaces","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3399715.3401952","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 1

Abstract

In this paper, we proposed the rotation based browsing method for video learning in personal training. SuppleView, which is flexible in respect of the user's physical position while viewing a video, enables coordinate translation free viewing between an observer and an actor. Previous work on video learning have not enough explored the limitation on the observation angle, although its angle effects for observer's comprehension and caused only in video learning not in observation with the actual trainer. The method solve this basic limitation by inferring the 3D pose of frames in a video. Based on those poses, we create an virtual agent with 3D model as an actor of movements, that is same with the movement in an original 2D video. The system transition for the two actors depends on the physical rotation of the user's head so that the angle of view for observing the actor also changes. Hence, the content rendering in proposed viewer could be provided to trainees as in kind-full form for their observation in the point of an observation angle of view. We report the method overview and our prototyping to show the proof of concept.

查看原文