International Journal of Pattern Recognition and Artificial Intelligence

A feature reduction technique is proposed for the hyperspectral (HS) data classification problem. The new features have been developed through a curve fitting step which fits specific rational function approximations to every spectral response curve (SRC) of HS image pixels. Then, the coefficients of the numerator and denominator polynomials of these fitted functions are considered as new extracted features. The method concentrates on the geometrical nature of SRCs and is utilizing the information that exists in sequence discipline — ordinance of reflectance coefficients in SRC — which has not been addressed by many other statistical analysis based methods. Maximum likelihood (ML) classification results show that the proposed method provides better classification accuracies compared to some basic and state-of-the-art feature extraction methods. Moreover, the proposed algorithm has the capability of being applied individually and simultaneously to all pixels of image.

The traditional vector-based classifiers, such as support vector machine (SVM) and twin support vector machine (TSVM), cannot handle tensor data directly and may not utilize the data informations effectively. In this paper, we propose a novel classifier based on tensor data, called twin bounded support tensor machine (TBSTM) which is an extension of twin bounded support vector machine (TBSVM). Similar to TBSVM, TBSTM gets two hyperplanes and obtains the solution by solving two quadratic programming problems (QPPs). The computational complexity of each QPPs is smaller than that of support tensor machine (STM). TBSTM not only retains the advantage of TBSVM, but also has its unique superior characteristics: (1) it makes full use of the structure information of data; (2) it has acceptable or better classification accuracy compared to STM, TBSVM and SVM; (3) the computational cost is basically less than STM; (4) it can deal with large data that TBSVM is not easy to achieve, especially for small-sample-size (S3) problems; (5) it adopts alternating successive over relaxation iteration (ASOR) method to solve optimization problems which accelerates the pace of training. Finally, we demonstrate the effectiveness and superiority by the experiments based on vector and tensor data.

There are a large number of methods for solving under-determined linear inverse problems. For large-scale optimization problem, many of them have very high time complexity. We propose a new method called two-stage sparse representation (TSSR) to tackle it. We decompose the representing space of signals into two parts”, the measurement dictionary and the sparsifying basis. The dictionary is designed to obey or nearly obey the sub-Gaussian distribution. The signals are then encoded on the dictionary to obtain the training and testing coefficients individually in the first stage. Then, we design the basis based on the training coefficients to approach an identity matrix, and we apply sparse coding to the testing coefficients over the basis in the second stage. We verify that the projection of testing coefficients onto the basis is a good approximation of the original signals onto the representing space. Since the projection is conducted on a much sparser space, the runtime is greatly reduced. For concrete realization, we provide an instance for the proposed TSSR. Experiments on four biometric databases show that TSSR is effective compared to several classical methods for solving linear inverse problem.

We describe a computationally efficient method to produce a specific Bayesian mixture of all the models in a finite set of feature-based models that assign a probability to the observed data set. Special attention is given to the bound on the regret of using the mixture instead of the best model in the set. It is proven theoretically and verified through synthetic data that this bound is relatively tight. Comparing the workload of the proposed method with the direct implementation of the Bayesian mixture shows an almost exponential improvement of computing time.

In computer vision, Local Binary Pattern (LBP) and Scale Invariant Feature Transform (SIFT) are two widely used local descriptors. In this paper, we propose to combine them effectively for scene categorization. First, LBP and SIFT features are regularly extracted from training images for constructing a LBP feature codebook and a SIFT feature codebook. Then, a two-dimensional table is created by combining the obtained codebooks. For creating a representation for an image, LBP and SIFT features extracted from the same positions of the image are encoded together based on sparse coding by using the two-dimensional table. After processing all features in the input image, we adopt spatial max pooling to determine its representation. Obtained image representations are forwarded to a Support Vector Machine classifier for categorization. In addition, in order to improve the scene categorization performance further, we propose a method to select correlated visual words from large codebooks for constructing the two-dimensional table. Finally, for evaluating the proposed method, extensive experiments are implemented on datasets Scene Categories 8, Scene Categories 15 and MIT 67 Indoor Scene. It is demonstrated that the proposed method is effective for scene categorization.

This paper extends the problem of abandoned/removed object classification in video surveillance to encompass the closely related (but to date ignored) problem of moved object classification. Existing abandoned/removed region classification techniques are shown to arbitrarily classify such localized object movements as either abandonment or removal. Our new approach correctly classifies abandoned and removed objects, and also correctly classified 90% of moved objects.

Co-saliency detection, an emerging research area in saliency detection, aims to extract the common saliency from the multi images. The extracted co-saliency map has been utilized in various applications, such as in co-segmentation, co-recognition and so on. With the rapid development of image acquisition technology, the original digital images are becoming more and more clearly. The existing co-saliency detection methods processing these images need enormous computer memory along with high computational complexity. These limitations made it hard to satisfy the demand of real-time user interaction. This paper proposes a fast co-saliency detection method based on the image block partition and sparse feature extraction method (BSFCoS). Firstly, the images are divided into several uniform blocks, and the low-level features are extracted from Lab and RGB color spaces. In order to maintain the characteristics of the original images and reduce the number of feature points as well as possible, Truncated Power for sparse principal components method are employed to extract sparse features. Furthermore, K-Means method is adopted to cluster the extracted sparse features, and calculate the three salient feature weights. Finally, the co-saliency map was acquired from the feature fusion of the saliency map for single image and multi images. The proposed method has been tested and simulated on two benchmark datasets: Co-saliency Pairs and CMU Cornell iCoseg datasets. Compared with the existing co-saliency methods, BSFCoS has a significant running time improvement in multi images processing while ensuring detection results. Lastly, the co-segmentation method based on BSFCoS is also given and has a better co-segmentation performance.

Brain-inspired algorithms such as convolutional neural network (CNN) have helped machine vision systems to achieve state-of-the-art performance for various tasks (e.g. image classification). However, CNNs mainly rely on local features (e.g. hierarchical features of points and angles from images), while important global structured features such as contour features are lost. Global understanding of natural objects is considered to be essential characteristics that the human visual system follows, and for developing human-like visual systems, the lost of consideration from this perspective may lead to inevitable failure on certain tasks. Experimental results have proved that well-trained CNN classifier cannot correctly distinguish fooling images (in which some local features from the natural images are chaotically distributed) from natural images. For example, a picture that is composed of yellow–black bars will be recognized as school bus with very high confidence by CNN. On the contrary, human visual system focuses on both the texture and contour features to form representation of images and would not mis-take them. In order to solve the upper problem, we propose a neural network model, named as histogram of oriented gradient (HOG) improved CNN (HCNN), that combines local and global features towards human-like classification based on CNN and HOG. The experimental results on MNIST datasets and part of ImageNet datasets show that HCNN outperforms traditional CNN for object classification with fooling images, which indicates the feasibility, accuracy and potential effectiveness of HCNN for solving image classification problem.

Sparse representation is a new hot technique in recent years. The two-phase test sample sparse representation method (TPTSSR) achieved an excellent performance in face recognition. In this paper, a kernel two-phase test sample sparse representation method (KTPTSSR) is proposed. Firstly, the input data are mapped into an implicit high-dimensional feature space by a nonlinear mapping function. Secondly, the data are analyzed by means of the TPTSSR method in the feature space. If an appropriate kernel function and the corresponding kernel parameter are selected, a test sample can be accurately represented as the linear combination of the training data with the same label information of the test sample. Therefore, the proposed method could have better recognition performance than TPTSSR. Experiments on the face databases demonstrate the effectiveness of our methods.

In this paper, we investigate the problem of personalized ranking from implicit feedback (PRIF). It is a more common scenario (e.g. purchase history, click log and page visitation) in recommender systems. The training data are only binary in these problems, reflecting the users’ actions or inactions. One shortcoming of previous PRIF algorithms is noise sensitivity: outliers in training data might bring significant fluctuations in the training process and lead to inaccuracy of the algorithm. In this paper, we propose two robust PRIF algorithms to solve the noise sensitivity problem of existing PRIF algorithms by using the pairwise sigmoid and pairwise fidelity loss functions. These two pairwise loss functions are flexible and can easily be adopted by popular collaborative filtering models such as the matrix factorization (MF) model and the K-nearest-neighbor (KNN) model. A learning process based on stochastic gradient descent with bootstrap sampling is utilized for the optimization. Experiments are conducted on practical datasets containing noisy data points or outliers. Results demonstrate that the proposed algorithms outperform several state-of-the-art one class collaborative filtering (OCCF) algorithms on both the MF and KNN models over different evaluation metrics.

In this paper, we propose a decomposition-based evolutionary algorithm with boundary search and archive for constrained multi-objective optimization problems (CMOPs), named CM2M. It decomposes a CMOP into a number of optimization subproblems and optimizes them simultaneously. Moreover, a novel constraint handling scheme based on the boundary search and archive is proposed. Each subproblem has one archive, including a subpopulation and a temporary register. Those individuals with better objective values and lower constraint violations are recorded in the subpopulation, while the temporary register consists of those individuals ever found before. To improve the efficiency of the algorithm, the boundary search method is designed. This method makes the feasible individuals with a higher probability to perform genetic operator with the infeasible individuals. Especially, when the constraints are active at the Pareto solutions, it can play its leading role. Compared with two algorithms, i.e. CMOEA/D-DE-CDP and Gary’s algorithm, on 18 CMOPs, the results show the effectiveness of the proposed constraint handling scheme.