hal-00203920, version 1
Accurate Object Detection with Deformable Shape Models Learnt from Images
Vittorio Ferrari
1, 2Frédéric Jurie 1, 2Cordelia Schmid 1, 2
Computer Vision and Pattern Recognition (CPRV '07) (2007) 1--8
Abstract: We present an object class detection approach which fully integrates the complementary strengths offered by shape matchers. Like an object detector, it can learn class models directly from images, and localize novel instances in the presence of intra-class variations, clutter, and scale changes. Like a shape matcher, it finds the accurate boundaries of the objects, rather than just their bounding-boxes. This is made possible by 1) a novel technique for learning a shape model of an object class given images of example instances; 2) the combination of Hough-style voting with a non-rigid point matching algorithm to localize the model in cluttered images. As demonstrated by an extensive evaluation, our method can localize object boundaries accurately, while needing no segmented examples for training (only bounding-boxes).
- 1: LEAR (INRIA Grenoble Rhône-Alpes / LJK Laboratoire Jean Kuntzmann)
- CNRS : FR71 – CNRS : UMR5527 – INRIA – Laboratoire Jean Kuntzmann – Université Joseph Fourier - Grenoble I – Institut National Polytechnique de Grenoble (INPG)
- 2: Laboratoire Jean Kuntzmann (LJK)
- CNRS : UMR5224 – Université Joseph Fourier - Grenoble I – Université Pierre Mendès-France - Grenoble II – Institut Polytechnique de Grenoble - Grenoble Institute of Technology
- Domain : Engineering Sciences/Signal and Image processing
Computer Science/Signal and Image Processing - Keywords : object detection – deformable shape models – shape learning
- hal-00203920, version 1
- http://hal.archives-ouvertes.fr/hal-00203920
- oai:hal.archives-ouvertes.fr:hal-00203920
- From: Véronique Rocher
- Submitted on: Monday, 21 January 2008 14:05:17
- Updated on: Friday, 3 December 2010 08:20:20
Associated documents
Export