K-087 Study of recognizing an unknown person action from motion capture data based on tensor decomposition
スポンサーリンク
概要
- 論文の詳細を見る
This paper applies an algorithm, based on Tensor Decomposition, to human action recognition application: by using motion database of people producing different actions, to recognize the action depicted in motion data generated by an unknown person. Human motion is the composite consequence of multiple elements, including the action performed and a motion signature that captures the distinctive pattern of movement of a particular individual. By performing decomposition, based on N- mode SVD (singular value decomposition), the algorithm analyzes motion data spanning multiple subjects performing different actions to extract the core tensors that govern the interaction between the mode matrices. The analysis yields a generative motion model that can recognize distinctive motion style of unknown person. The effectiveness of applying the tensor decomposition approach to our purpose was confirmed by recognizing novel walking motions for a person by using motion database.
- FIT(電子情報通信学会・情報処理学会)推進委員会の論文
- 2005-08-22
著者
関連論文
- J-021 The Prototype system of Prompter communication with handwriting and drawing functions
- K_089 Analysis of Communication Prompted by Black "●" Communication System : Effects of Emotional Emphasis in Comparison with E-mails and Messengers
- I_053 Object Detection by Using Marker Assistance
- O-005 Ecological-value Based End-of-life Product Collecting Network Re-design and Optimization
- H-029 Study of Optimizing the Parameters in the Weighted Histogram Equalization for Stereoscopic Enhancement
- I-059 Image-based Assistance to Fire Extinguishing System with Mote
- H-003 Comparison of Several Shape Descriptors in Order to Perform Car Recognition from a Moving Camera
- E-033 Proposed Method for OMR of Square Notation Manuscripts
- J-024 Tensor Decomposition Framework For Recognizing an Unknown Person's Action From A Video Sequence Using Image Features
- J-050 Comparative Study of Non-rigid Objects' Animations By Spring models and Boundary Element Method
- I_074 Study of Color Contrast Enhancement Based on Wavelet Transform
- K-087 Study of recognizing an unknown person action from motion capture data based on tensor decomposition
- K-035 A Frame Work for Analysis and Synthesis of Actor's Stylistic Action
- I-066 Low Level Feature Detection based on Modified Ransac