Paper Info
Title
Discriminative Patch Selection using Combinatorial and Statistical Models for Patch-Based Object Recognition
Abstract
In an object recognition task where an image is represented as a constellation of image patches, often many patches correspond to the cluttered background. If such patches are used for object class recognition, they will adversely affect the recognition rate. In this paper, we present a two stage method for selecting image patches which characterize the target object class and are capable of discriminating between the positive images containing the target objects and the complementary negative images. The first stage selection is done using a novel combinatorial optimization formulation on a weighted multipartite graph representing similarities between images patches across different instances of the target object. The following stage is a statistical method for selecting those images patches from the positive images which, when used individually, have the power of discriminating between the positive and negative images in the evaluation data. The individual methods have a performance competitive with the state of the art methods on a popular benchmark data set and their sequential combination consistently outperforms the individual methods and most of the other known methods while approaching the best known results.
Year
DOI
Venue
2006
10.1109/CVPRW.2006.66
CVPR Workshops
Field
DocType
Volume
Computer vision,Object detection,Viola–Jones object detection framework,3D single-object recognition,Three-dimensional face recognition,Object-class detection,Pattern recognition,Computer science,Feature extraction,Feature (machine learning),Artificial intelligence,Cognitive neuroscience of visual object recognition
Conference
2006
Issue
ISSN
ISBN
1
2160-7508
0-7695-2646-2
Citations 
PageRank 
References 
2
0.48
16
Authors
5
Name
Order
Citations
PageRank
Akshay Vashist117612.64
Zhipeng Zhao2384.08
Ahmed Elgammal32553168.71
Ilya Muchnik432347.03
casimir a kulikowski5616299.37