Open Access System for Information Sharing

Login Library

 

Thesis
Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

A Comparative Assessment on the Practicality of the Signal-based Human Motion Recognition Methods

Title
A Comparative Assessment on the Practicality of the Signal-based Human Motion Recognition Methods
Authors
정하영
Date Issued
2018
Publisher
포항공과대학교
Abstract
신호 기반 모션 인식 기법은 최근 학계에서 많은 주목을 받았음에도 불구하고 시장으로 뻗어나가는 데에는 실패하였다. 우리는 신호 기반 모션 인식 기법들이 가진 비현실적인 요구사항의 한계를 지적하며 일곱 개의 실용성 기준 - 입상도(granularity), 안정성(stability), 강건성(robustness), 응용성(applicability), 배치성(deployability), 효율성(efficiency) - 을 제안하였다. 이에 기초하여 기존에 발표된 시그널 기반 모션 기법들을 비교 분석 및 평가하였고 이를 통해 얻은 발견과 최신 동향 및 장래의 연구 방향을 제시하였다. 실용성을 평가하는 기준을 통해 미래 시그널 기반 모션 인식 연구 방향이 연구를 위한 연구를 넘어서 실질적으로 보급되는 방향으로 나아갈 것을 기대한다.
Signal-based motion recognition schemes have received growing attention in the academia, but they failed to be adopted by the industry despite their flourish. In this thesis, we address the impractical requirements and conditions of signal-based human motion recognition methods. We propose seven practical facets - granularity, stability, robustness, applicability, deployability, and efficiency - that extensively evaluate the practicality of a scheme. We perform evaluation on the previous signal-based human motion recognition research based on the seven facets and unfold the findings we have earned through the analysis in addition to the emerging trends in the field and prospect on future research directions. We believe our work may serve as a standard for future signal-based human motion recognition research to assess the practicality of their schemes.
URI
http://postech.dcollection.net/common/orgView/200000006967
https://oasis.postech.ac.kr/handle/2014.oak/93577
Article Type
Thesis
Files in This Item:
There are no files associated with this item.

qr_code

  • mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Views & Downloads

Browse