International Journal of Wavelets, Multiresolution and Information Processing

The accuracy of learning resource recommendation is crucial to realizing precise teaching and personalized learning. We propose a novel collaborative filtering recommendation algorithm based on the student’s online learning sequential behavior to improve the accuracy of learning resources recommendation. First, we extract the student’s learning events from his/her online learning process. Then each student’s learning events are selected as the basic analysis unit to extract the feature sequential behavior sequence that represents the student’s learning behavioral characteristics. Then the extracted feature sequential behavior sequence generates the student’s feature vector. Moreover, we improve the H-$K$ clustering algorithm that clusters the students who have similar learning behavior. Finally, we recommend learning resources to the students combine similarity user clusters with the traditional collaborative filtering algorithm based on user. The experiment shows that the proposed algorithm improved the accuracy rate by 110% and recall rate by 40% compared with the traditional user-based collaborative filtering algorithm.

In video surveillance, the captured face images are usually suffered from low-resolution (LR), besides, not all the probe images have mates in the gallery under the premise that only a single frontal high-resolution (HR) face image per subject. To address this problem, a novel face recognition framework called recursive label propagation based on statistical classification (ReLPBSC) has been proposed in this paper. Firstly, we employ VGG to extract robust discriminative feature vectors to represent each face. Then we select the corresponding LR face in the probe for each HR gallery face by similarity. Based on the picked HR–LR pairs, ReLPBSC is implemented for recognition. The main contributions of the proposed approach are as follows: (i) Inspired by substantial achievements of deep learning methods, VGG is adopted to achieve discriminative representation for LR faces to avoid the super-resolution steps; (ii) the accepted and rejected threshold parameters, which are not fixed in face recognition, can be achieved with ReLPBSC adaptively; (iii) the unreliable subjects never enrolled in the gallery can be rejected automatically with designed methods. Experimental results in $16\times 16$ pixels resolution show that the proposed method can achieve 86.64% recall rate while keeping 100% precision.

For the object loss problem in the tracking process caused by illumination, occlusion, pose variation, and motion blur, the tracking method based on dual fuzzy low-rank approximation in a particle filter framework is proposed in this paper. Firstly, multiple constraint regions are built to filter insignificant samples, and more distinguished candidate samples are selected. Secondly, dual fuzzy observation function of each candidate sample is created based on the designed low-rank approximation representations of object and background. Then the generalized tracking results are obtained by computing membership degrees of dual fuzzy observation functions. Finally, based on the spatial coherency principle, the final tracking result is determined from the generalized results by measuring similarities of consecutive objects. The proposed method shows good performance as compared with several state-of-the-art trackers on challenging benchmark sequences.

Radon transform is not only robust to noise, but also independent on the calculation of pattern centroid. In this paper, Radon–Mellin transform (RMT), which is a combination of Radon transform and Mellin transform, is proposed to extract invariant features. RMT converts any object into a closed curve. Radon–Fourier descriptor (RFD) is derived by applying Fourier descriptor to the obtained closed curve. The obtained RFD is invariant to scaling and rotation. (Generic) R-transform and some other Radon-based methods can be viewed as special cases of the proposed method. Experiments are conducted on some binary images and gray images.

To improve the performance of smoke recognition, it is important to extract representative and discriminative features for smoke. We propose a learning based method to automatically extract discriminative features from raw pixels. First, patch-based local differences rather than single-pixel-based differences are computed. Then a sampling strategy is learnt to extract dominant local differences. Third, the sampled differences are used to learn a codebook, through which differences are projected into a discriminative feature space. Thus, feature matrices consisting of discriminative differences can be obtained according to the codebook. At last, the feature maps are encoded without quantization, thus final feature is computed. Experimental results demonstrate that the proposed method achieves better performance than some existing handcrafted and learning based feature extraction methods.

Nonnegative matrix factorization (NMF) is a promising method to represent facial images using nonnegative features under a low-rank nonnegative basis-image matrix. The facial images usually reside on a low-dimensional manifold due to the variations of illumination, pose and facial expression. However, NMF has no ability to uncover the manifold structure of data embedded in a high-dimensional Euclidean space, while the manifold structure contains both local and nonlocal intrinsic features. These two kinds of features are of benefit to class discrimination. To enhance the discriminative power of NMF, this paper proposes a novel NMF algorithm with manifold structure (Mani-NMF). Two quantities related to adjacent graph and non-adjacent graph are incorporated into the objective function, which will be minimized by solving two convex suboptimization problems. Based on the gradient descent method and auxiliary function technique, we acquire the update rules of Mani-NMF and theoretically prove the convergence of the proposed Mani-NMF algorithm. Three publicly available face databases, Yale, pain expression and CMU databases, are selected for evaluations. Experiments results show that our algorithm achieves a better performance than some state-of-the-art algorithms.

Biologically inspired model (BIM) for image recognition is a robust computational architecture, which has attracted widespread attention. BIM can be described as a four-layer structure based on the mechanisms of the visual cortex. Although the performance of BIM for image recognition is robust, it takes the randomly selected ways for the patch selection, which is sightless, and results in heavy computing burden. To address this issue, we propose a novel patch selection method with oriented Gaussian–Hermite moment (PSGHM), and we enhanced the BIM based on the proposed PSGHM, named as PBIM. In contrast to the conventional BIM which adopts the random method to select patches within the feature representation layers processed by multi-scale Gabor filter banks, the proposed PBIM takes the PSGHM way to extract a small number of representation features while offering promising distinctiveness. To show the effectiveness of the proposed PBIM, experimental studies on object categorization are conducted on the CalTech05, TU Darmstadt (TUD) and GRAZ01 databases. Experimental results demonstrate that the performance of PBIM is a significant improvement on that of the conventional BIM.

Unmanned aerial vehicles (UAVs) equipped with high definition (HD) cameras can obtain a large number of detailed inspection images. The insulator is an indispensable component in the transmission lines. Detecting insulator in image video quickly and accurately can provide a reliable basis for the ranging and the obstacle avoidance flight of UAV close to the tower and transmission line. At the same time, the insulator is a serious threat to the safety of the power grid due to the multiple faults of the insulator, and the computer technology should be fully utilized to diagnose the fault. Detection of the insulator images with the complex aerial background is implemented by constructing a convolutional neural network (CNN), which has the classic architecture of five modules of convolution and pooling, two modules of fully connected layers. In this paper, we propose a recognition algorithm for explosion fault based on saliency detection, which uses the trained network model to extract the features. Then, we put the saliency maps into a self-organizing feature map (SOM) network and build the mathematical module via super pixel segmentation, contour detection and other image processing methods. The test shows that the algorithm can reduce the error that may be caused by manual analysis. It also demonstrates that the detection of the insulator and the recognition of explosion fault can effectively improve the efficiency and intelligence level.

Automatic human action recognition is a core functionality of systems for video surveillance and human object interaction. In the whole recognition system, feature description and encoding represent two crucial key steps. In order to construct a powerful action recognition framework, it is important that the two steps must provide reliable performance. In this paper, we proposed a new human action feature descriptor which is called spatio-temporal histograms of gradients (SPHOG). SPHOG is based on the spatial and temporal derivation signal, which extracts the gradient changes between consecutive frames. Compared to the traditional descriptors histograms of optical flow, our proposed SPHOG costs less computation resource. In order to incorporate the distribution information of local descriptors into Vector of Locally Aggregated Descriptors (VLAD), which is a popular encoding approach for Bag-of-Feature representation, a Gaussian kernel is implanted to compute the weighted distance histograms of local descriptors. By doing this, the encoding schema for bag-of-feature (BOF) representation is more effective. We validated our proposed algorithm for human action recognition on three public available datasets KTH, UCF Sports and HMDB51. The evaluation experiment results indicate that the proposed descriptor and encoding method can improve the efficiency of human action recognition and the recognition accuracy.

The high dimensionality and heterogeneity of the hyperspectral image (HSI) make a challenge to the application of machine learning methods, such as sparse subspace clustering (SSC). SSC is designed to represent data as an union of affine subspaces, while it cannot capture the latent structure of the given data. In Mosers theory, the distribution can represent the intrinsic structure efficiently. Hence, we propose a novel approach called spatial distribution preserving-based sparse subspace clustering (SSC-SDP) in this paper for HSI data, which can help sparse representation preserve the underlying manifold structure. Specifically, the density constraint is added by minimizing the inconsistency of the densities estimated in the HSI data and the corresponding sparse coefficient matrix. In addition, we incorporate spatial information into the density estimation of the original data, and the optimization solution based on alternating direction method of multipliers (ADMM) is devised. Three HSI data sets are conducted to evaluate the performance of our SSC-SDP compared with other state-of-art algorithms.

With the extensive attention and research of the scholars in deep learning, the convolutional restricted Boltzmann machine (CRBM) model based on restricted Boltzmann machine (RBM) is widely used in image recognition, speech recognition, etc. However, time consuming training still seems to be an unneglectable issue. To solve this problem, this paper mainly uses optimized parallel CRBM based on Spark, and proposes a parallel comparison divergence algorithm based on Spark and uses it to train the CRBM model to improve the training speed. The experiments show that the method is faster than traditional sequential algorithm. We train the CRBM with the method and apply it to breast X-ray image classification. The experiments show that it can improve the precision and the speed of training compared with traditional algorithm.