2024-03-29T02:18:57Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:000948442023-04-27T10:00:04Z01164:04619:06988:07247
個人差と個人内変動を分離した統計的顔形状モデルStatistical face shape model separating inter-individual variation from intra-individual variationjpnhttp://id.nii.ac.jp/1001/00094825/Technical Reporthttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=94844&item_no=1&attribute_id=1&file_no=1Copyright (c) 2013 by the Institute of Electronics, Information and Communication EngineersThis SIG report is only available to those in membership of the SIG.(株)豊田中央研究所小島, 真一近年 Point Distribution Model (PDM) と呼ばれる顔形状を数十点の点群で表現するモデリング手法が顔画像処理の分野で使われている.PDM は事前に用意した顔形状集合から主成分分析で抽出した少数基底ベクトルの重み付き線形和で表現するのが特徴だが,個人差を表現する基底と個人内変動を表現する基底が分離できていないという問題があった.本研究では上記問題を解決した,個人差を表現する基底と個人内変動を表現する基底を線形分離した統計的顔形状モデルの構成法について述べる.Point Distribution Model is a shape model of Active Appearance Models used in face image processing, that is represented by a weighted linear sum of a few basis vectors that were extracted with principal component analysis from the collected face geometry data. The problem of that model is that inter-individual variation and intra-individual variation were not separated. In this paper, we describe a composing method of statistical face shape model that is linearly separated inter-individual variation and intra-individual variation.AA11131797研究報告コンピュータビジョンとイメージメディア(CVIM)2013-CVIM-1883162013-08-262013-08-19