Multi-view Object Category Detection

• Autorzy: Szumilas L. , Wildenauer H.
• Języki publikacji: EN

Abstrakty

EN

This paper proposes a novel method for object category detection in images, which is based on spatial configuration of local shape features. We show how simple, redundant edge based features overcome the problem of edge fragmentation while the efficient use of geometrically related feature pairs allows us to construct a robust object shape matcher, invariant to translation, scale and rotation. These prerequisites are used for weakly supervised learning of object models as well as object class detection. The object models employing pairwise combination of redundant shape features exhibit remarkably accurate localization of similar objects even in the presence of clutter and moderate view point changes which is further exploited for model building, object detection and recognition.

Słowa kluczowe

EN

robotics; object identification; European project

PL

robotyka; identyfikacja obiektów; projekty europejskie

• Wydawca: Oficyna Wydawnicza Politechniki Warszawskiej
• Czasopismo: Prace Naukowe Politechniki Warszawskiej. Elektronika
• Rocznik: 2010
• Tom: z. 175, t. 2
• Strony: 677--690
• Opis fizyczny: Bibliogr. 16 poz., tab., rys., wykr.

Twórcy

autor

Szumilas L.

autor

Wildenauer H.

• Industrial Research Institute for Automation and Measurements, Al. Jerozolimskie 202, 02-486 Warsaw, Poland,

Bibliografia

• [1] http://www.pascal-network.org/challengcs/VOC.
• [2] Dana H. Ballard. Generalizing the hough transform to detect arbitrary shapes. Pattern Recognition, 2(13), 1981.
• [3] Alexander C. Berg, Tamara L. Berg, and Kitandra Malik. Shape matching and object recognition using low distortion correspondences. In CVPR, pages 26-33, 2005.
• [4] David Crandall, Pedro Felzenszwalb, and Daniel Huttenlocher. Spatial priors for part-based recognition using statistical models. In CVPR, pages 10-17, 2005.
• [5] P. Felzenszwalb, D. Mcallester, and D. Ramanan. A discriminatively trained, multiscale, deformable part model. In CVPR, June 2008.
• [6] Vittorio Ferrari, Loic Fevrier, Frederic Jurie, and Cordelia Schmid. Groups of adjacent contour segments for object detection. T-PAMI, 30(1), 2008.
• [7] Tae-Kyun Kim and Roberto Cipolla. Mcboost: Multiple classifier boosting for perceptual co-clustering of images and visual features. In NIPS, pages 841-856, 2008.
• [8] Ivan Laptev. Improving object detection with boosted histograms. Image and Vision mComputing, 2008.
• [9] Bastian Leibe, Ales Leonardis, and Bernt Schiele. Robust object detection with inter-leaved categorization and segmentation. IJCV, 77(1-3):259-289, 2008.
• [10] Marius Leordeanu and Martial Hebert. A spectral technique for correspondence problems using pairwise constraints. In ICCV, volume 2, pages 1482-1489, October 2005.
• [11] Marius Leordeanu and Martial Hebert. Beyond local appearance: Category recognition from pairwise interactions of simple features. In CVPR, pages 1-8, 2007.
• [12] Andreas Opelt, Axel Pinz, and Andrew Zisserman. A boundary-fragment-model for object detection. In ECCV, pages 575-588, 2006.
• [13] Pietro Perona, Robert Fergus and Andrew Zisserman. A sparse object category model for efficient learning and exhaustive recognition. In CVPR, pages 380-387, 2005.
• [14] Jamie Shotton, Andrew Blake, and Robert Cipolla. Multiscale categorical object recognition using contour fragments. T-PAMI, 30(7): 1270-1281, 2008.
• [15] Vladimir Vapnik, Steven E. Golowich, and Alex J. Smola. Support vector method for function approximation, regression estimation and signal processing. In NIPS, pages 281-287, 1996.
• [16] John Win and Nebojsa Jojic. Locus: learning object classes with unsupervised segmentation. In ICCV, pages 756-763, 2005.