Search

The main objective of this research paper is to design a system which would generate multimodal, nonparametric Bayesian model, and multilayered probability latent semantic analysis (pLSA)-based visual dictionary (BM-MpLSA). Advancement in technology and the exuberance of sports lovers have necessitated a requirement for automatic action recognition in the live video seed of sports. The fundamental requirement for such model is the creation of visual dictionary for each sports domain. This multimodal nonparametric model has two novel co-occurrence matrix creation—one for image feature vector and the other for textual entities. This matrix provides a basic scaling parameter for the unobserved random variables, and it is an extension of multilayered pLSA-based visual dictionary creation. This paper precisely concentrates on the creation of visual dictionary for Basketball. From the sports event images, the feature vector extracted is modified as SIFT and MPEG 7's-based dominant color, color layout, scalable color and edge histograms. After quantization and analysis of these vector values, the visual vocabulary would be created by integrating them into the domain specific visual ontology for semantic understanding. The accuracy rate of this work is compared with respect to the action held on image based on performance. [ABSTRACT FROM AUTHOR]

This paper presents an efficient word spotting system applied to handwritten Arabic documents, where images are represented with bag-of-visual-SIFT descriptors and a sliding window approach is used to locate the regions that are most similar to the query by following the query-by-example paragon. First, a pre-processing step is used to produce a better representation of the most informative features. Secondly, a region-based framework is deployed to represent each local region by a bag-of-visual-SIFT descriptors. Afterward, some experiments are in order to demonstrate the codebook size influence on the efficiency of the system, by analyzing the curse of dimensionality curve. In the end, to measure the similarity score, a floating distance based on the descriptor's number for each query is adopted. The experimental results prove the efficiency of the proposed processing steps in the word spotting system. [ABSTRACT FROM AUTHOR]

Recent years, face recognition based on video has been concerned by more and more persons. At the same time, bag-of-visual words (BoWs) representation has been successfully applied in image retrieval and object recognition recently. In this paper, a video-based face recognition approach which uses visual words Is proposed. In classc visual words, scale invariant feature transform (SIFT) descriptors of an image are firstly extracted on interest points detected by difference of Gaussian (DoG), then k-means-based visual vocabulary generation is applied to replace these descriptors wtth the indexes of the closet visual words. However, in facial images, SIFT descriptors are not good enough due to facial pose distortion, facial expression and lighting condition variation. In this paper, we use Affine-SIFT (ASIFT) descriptors as facial image representation. Experimental results on Yale and ORL Database suggest that proposed method can achieve lower error rates in face recognition task. [ABSTRACT FROM AUTHOR]

In this article, we propose a novel Hamming embedding kernel with informative bag-of-visual words to address two main problems existing in traditional BoW approaches for video semantic indexing. First, Hamming embedding is employed to alleviate the information loss caused by SIFT quantization. The Hamming distances between keypoints in the same cell are calculated and integrated into the SVM kernel to better discriminate different image samples. Second, to highlight the concept-specific visual information, we propose to weight the visual words according to their informativeness for detecting specific concepts. We show that our proposed kernels can significantly improve the performance of concept detection. [ABSTRACT FROM AUTHOR]

Abstract: Given the explosive growth of the Web images, image search plays an increasingly important role in our daily lives. The visual representation of image is the fundamental factor to the quality of content-based image search. Recently, bag-of-visual word model has been widely used for image representation and has demonstrated promising performance in many applications. In the bag-of-visual-word model, the codebook/visual vocabulary plays a crucial role. The conventional codebook, generated via unsupervised clustering approaches, does not embed the labeling information of images and therefore has less discriminative ability. Although some research has been conducted to construct codebooks with the labeling information considered, very few attempts have been made to exploit manifold geometry of the local feature space to improve codebook discriminative ability. In this paper, we propose a novel discriminative codebook learning method by introducing the subspace learning in codebook construction and leveraging its power to find a contextual local descriptor subspace to capture the discriminative information. The discriminative codebook construction and contextual subspace learning are formulated as an optimization problem and can be learned simultaneously. The effectiveness of the proposed method is evaluated through visual reranking experiments conducted on two real Web image search datasets. [Copyright &y& Elsevier]