In this paper, we consider the problem of approximate nearest neighbor (ANN) retrieval with the method of sparsecoding, which is a promising tool of discovering compact representation of high-dimensional data. A new ...
详细信息
In this paper, we consider the problem of approximate nearest neighbor (ANN) retrieval with the method of sparsecoding, which is a promising tool of discovering compact representation of high-dimensional data. A new study, exploiting the indices of the active set of sparse coded data as retrieval code, exhibits an appealing ANN route. Here our work aims to enhance the method via considering its shortages of the local structure of the data. Our primary innovation is two-fold: We introduce the graph Laplacian regularization to preserve the local structure of the original data into reduced space, which is indeed beneficial to ANN. And we impose non-negativity constraints such that the retrieval code can more effectively reflect the neighborhood relation, thereby cutting down on unnecessary hash collision. To this end, we learn an incoherent dictionary with both rules via a novel formulation of sparsecoding. The resulting optimization problem can be provided with an available solution by the widely used iterative scheme, where we resort to the feature-sign search algorithm in the sparsecoding step and exploit the method that uses a Lagrange dual for dictionary learning step. Experimental results on publicly available image data sets manifest that the rules are positive to promote the classification and ANN accuracies. Compared with several state-of-the-art ANN techniques, our methods can achieve an interesting improvement in retrieval accuracy. (C) 2017 Elsevier Ltd. All rights reserved.
Among the existing hashing methods, the Self-taught hashing (STH) is regarded as the state-of-the-art work. However, it still suffers the problem of semantic loss, which mainly comes from the fact that the original op...
详细信息
Among the existing hashing methods, the Self-taught hashing (STH) is regarded as the state-of-the-art work. However, it still suffers the problem of semantic loss, which mainly comes from the fact that the original optimization objective of in-sample data is NP-hard and therefore is compromised into the combination of Laplacian Eigenmaps (LE) and binarization. Obviously, the shape associated with the embedding of LE is quite dissimilar to that of binary code. As a result, binarization of the LE embedding readily leads to significant semantic loss. To overcome this drawback, we combine the constrained nonnegative sparse coding and the Support Vector Machine (SVM) to propose a new hashing method, called nonnegative sparse coding induced hashing (NSCIH). Here, nonnegative sparse coding is exploited for seeking a better intermediate representation, which can make sure that the binarization can be smoothly conducted. In addition, we build an image copy detection scheme based on the proposed hashing methods. The extensive experiments show that the NSCIH is superior to the state-of-the-art hashing methods. At the same time, this copy detection scheme can be used for performing copy detection over very large image database. (C) 2012 Elsevier B.V. All rights reserved.
In this study, we improved the visualization method for attack patterns in team sports that was proposed in our previous work to visualize not only patterns of offense, but also patterns of defense. In the improved me...
详细信息
ISBN:
(纸本)9781538626337
In this study, we improved the visualization method for attack patterns in team sports that was proposed in our previous work to visualize not only patterns of offense, but also patterns of defense. In the improved method, dictionary learning for nonnegative sparse coding is performed to obtain a dictionary of base player densities, which can be considered as components of player formations for both teams. Next, nonnegative sparse coding is utilized to evaluate coefficient vectors as features in corresponding scenes. Finally, co-occurrence analysis based on mutual self-information is utilized to construct a co-occurrence network for shooting and formation components for both teams. For validation, the improved method was applied to player position data from five matches from the 2011 J. League.
This study proposes a novel non-negative matrix factorisation (NMF) variant L-1/2-NMF after visualisation and analysis of the process of target recognition via NMF for synthetic aperture radar (SAR) images. NMF has be...
详细信息
This study proposes a novel non-negative matrix factorisation (NMF) variant L-1/2-NMF after visualisation and analysis of the process of target recognition via NMF for synthetic aperture radar (SAR) images. NMF has been applied to obtain pattern feature in SAR images. This study considers the intrinsic character and the physical meaning of NMF feature when applied for SAR automatic target recognition. At the base of obtaining the linear relationship between the sample to be recognised and the train samples, the whole recognition process is detailed and vividly visualised. Meanwhile, lots of researches have been done to improve NMF methods by enforcing sparse constraint with L-1-norm, such as non-negative sparsecoding (NNSC), local NMF and sparse NMF. Compared with L-1-norm, L-1/2-norm has been shown to have a more natural sparseness. In this study, a novel variant of NMF with L-1/2 constraint, called L-1/2-NMF is proposed, and is carried out a thorough study by applying it in SAR target recognition. Experimental results on MSTAR public database show that both the basis and coding matrices obtained by L-1/2-NMF have higher sparseness than those obtained by NMF, NNSC and NMF with sparseness constraints (NMFsc). The recognition results demonstrate that the L-1/2-NMF outperforms NNSC, NMFsc and non-smooth NMF.
In this article, we present a novel approach to segment discriminative patches in human activity videos. First, we adopt the spatio-temporal interest points (STIPs) to represent significant motion patterns in the vide...
详细信息
In this article, we present a novel approach to segment discriminative patches in human activity videos. First, we adopt the spatio-temporal interest points (STIPs) to represent significant motion patterns in the video sequence. Then, nonnegative sparse coding is exploited to generate a sparse representation of each STIP descriptor. We construct the feature vector for each video by applying a two-stage sum-pooling and l(2)-normalization operation. After training a multi-class classifier through the error-correcting code SVM, the discriminative portion of each video is determined as the patch that has the highest confidence while also being correctly classified according to the video category. Experimental results show that the video patches extracted by our method are more separable, while preserving the perceptually relevant portion of each activity.
We present a novel perspective on shape characterization using the screened Poisson equation. We discuss that the effect of the screening parameter is a change of measure of the underlying metric space. Screening also...
详细信息
We present a novel perspective on shape characterization using the screened Poisson equation. We discuss that the effect of the screening parameter is a change of measure of the underlying metric space. Screening also indicates a conditioned random walker biased by the choice of measure. A continuum of shape fields is created by varying the screening parameter or, equivalently, the bias of the random walker. In addition to creating a regional encoding of the diffusion with a different bias, we further break down the influence of boundary interactions by considering a number of independent random walks, each emanating from a certain boundary point, whose superposition yields the screened Poisson field. Probing the screened Poisson equation from these two complementary perspectives leads to a high-dimensional hyperfield: a rich characterization of the shape that encodes global, local, interior, and boundary interactions. To extract particular shape information as needed in a compact way from the hyperfield, we apply various decompositions either to unveil parts of a shape or parts of a boundary or to create consistent mappings. The latter technique involves lower-dimensional embeddings, which we call screened Poisson encoding maps (SPEM). The expressive power of the SPEM is demonstrated via illustrative experiments as well as a quantitative shape retrieval experiment over a public benchmark database on which the SPEM method shows a high-ranking performance among the existing state-of-the-art shape retrieval methods.
So far, the most popular method for object & scene categorization (such as Vector Quantization (VQ), sparsecoding (SC)) transforms low-level descriptors (usually SIFT descriptors) into mid-level representations w...
详细信息
ISBN:
(纸本)9780819484185
So far, the most popular method for object & scene categorization (such as Vector Quantization (VQ), sparsecoding (SC)) transforms low-level descriptors (usually SIFT descriptors) into mid-level representations with more meaningful information. These methods have two key steps: (1) the building dictionary step, which provides a mechanism to map low-level descriptors into mid-level representations. (2) The coding step, which implements the map from low-level descriptors to mid-level representation for each image by the dictionary. In this paper, we proposed to use a stable and efficient nonnegative sparse coding (SENSC) algorithm for building dictionary and coding each image with it to develop an extension of Spatial Pyramid Matching (SPM) method. We also compare SENSC with SC (state-of-the-art performance) method and VQ method, analysis the drawbacks of SC and VQ for building dictionary, and show SENSC algorithm's performance. According to the experiments on three benchmarks (Caltech101, scene, and events), the method we proposed has shown a better performance than SC and VQ methods.
暂无评论