Presentation 1994/1/21
Learning and Recognition of 3D Object from Appearance
Hiroshi Murase, Shree Nayar,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We address the problem of leaning object models for recognition and pose estimation.We formulate the recognition problem as one of matching visual appearance rather than shape.A new compact image representation called parametric eigenspace is proposed.The image set is compressed to obtain a low-dimensional subspace in which the object is represented as a hypersurface parametrized by pose and illumination.The recognition system projects the image onto the eigenspace.The object is recognized based on the hypersurface it lies on.The position of the projection on the hypersurface determines the object′s pose.We have conducted experiments using s everal objects with complex appearance characteristics.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Object Recognition / Learning / Eigenvector / Pose Estimation / Principal Component Analysis
Paper # PRU93-120
Date of Issue

Conference Information
Committee PRU
Conference Date 1994/1/21(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Understanding (PRU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Learning and Recognition of 3D Object from Appearance
Sub Title (in English)
Keyword(1) Object Recognition
Keyword(2) Learning
Keyword(3) Eigenvector
Keyword(4) Pose Estimation
Keyword(5) Principal Component Analysis
1st Author's Name Hiroshi Murase
1st Author's Affiliation Basic Research Laboratory,NTT()
2nd Author's Name Shree Nayar
2nd Author's Affiliation Department of Computer Science,Columbia University
Date 1994/1/21
Paper # PRU93-120
Volume (vol) vol.93
Number (no) 431
Page pp.pp.-
#Pages 8
Date of Issue