J-GLOBAL ID:200904056165102210  Research Project code:0250004348 Update date:Jan. 17, 2003

Extraction and classification of context by self-organization of spatio-temporal information

文脈主導型,認識・判断・行動機能を有した工学システム実現に関する研究 文脈事象の想起・学習のソフトウェアモデルに関する研究 時空間情報の自己組織化による文脈の抽出,判定
Study period:1998 - 2002
Organization (1):
Investigating Researcher (1):
Research overview:
Feature extraction from dynamically changing image scene is an inherently difficult research task due to huge amount of information. The present research extends self-organization for static patterns to that for dynamic patterns. This develops technologies for constructing a context by integrating similar sequences, judging to which context a given sequence belongs, and judging a novel sequence as a new context.
Keywords (6):
neural network ,  learning and self-organization ,  gesture recognition ,  mobile robot ,  information integration ,  rule discovery
Project name: Research on dynamical memory system for realizing context dependent recognition, judgement and behavioral functions
Project Organization (1):
  • Saitama University
Researcher representative of the project  (1):
Research program: Special Coordination Funds for Promoting Science and Technology
Ministry with control over the research :
Ministry of Education, Culture, Sports, Science and Technology
Research budget: 2002: \3,505,000

Return to Previous Page