Real-time synthesis of a humanlike agent in response to the user's moving image

Osamu Hasegawa*, Chil Woo Lee, Wiwat Wongwarawipat, Mitsuru Ishizuka

*この研究の対応する著者

研究成果: Article査読

1 被引用数 (Scopus)

抄録

This paper describes a parallel computing system and a software algorithm for real-time interaction between a human user and a synthesized, humanlike, moving image. The realistic humanlike agent on a monitor can recognize the palm position and finger motion (finger sign) of the user. She/he then tracks and gazes at the hand position to change her/his facial expression in response to the finger sign in real time. This visual software agent (VSA) is expected to play an important role in building an advanced human interface. We regard this type of interactive agent as a visual software robot. To achieve real-time image recognition and image synthesis, we have developed a parallel visual computer system the transputer network with visual interface to transputers (TN-VIT). The imagesynthesis speed of the TN-VIT is about 24 frames/s, including finger-sign recognition. Some samples of synthesized images and experimental results are shown.

本文言語English
ページ(範囲)122-130
ページ数9
ジャーナルMachine Vision and Applications
8
2
DOI
出版ステータスPublished - 1995 3
外部発表はい

ASJC Scopus subject areas

  • 電子工学および電気工学
  • 制御およびシステム工学
  • 人間とコンピュータの相互作用
  • コンピュータ ビジョンおよびパターン認識
  • 人工知能

フィンガープリント

「Real-time synthesis of a humanlike agent in response to the user's moving image」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル