Low-dimensional Discriminative Features Research Articles

Euler Representation-Based Structural Balance Discriminant Projection for Machinery Fault Diagnosis

Fault diagnosis methods are usually sensitive to outliers and it is difficult to obtain and balance global and local discriminant information, which leads to poor separation between classes of low-dimensional discriminant features. For this problem, we propose an Euler representation-based structural balance discriminant projection (ESBDP) algorithm for rotating machine fault diagnosis. First, the method maps the high-dimensional fault features into the Euler representation space through the cosine metric to expand the differences between heterogeneous fault samples while reducing the impact on outliers. Then, four objective functions with different structure and class information are constructed in this space. On the basis of fully mining the geometric structure information of fault data, the local intra-class aggregation and global inter-class separability of the low-dimensional discriminative features are further improved. Finally, we provide an adaptive balance strategy for constructing a unified optimization model of ESBDP, which achieves the elastic balance between global and local features in the projection subspace. The diagnosis performance of the ESBDP algorithm is explored by two machinery fault cases of bearing and gearbox. Encouraging experimental results show that the algorithm can capture effective fault discriminative features and can improve the accuracy of fault diagnosis.

Robust Low Rank and Sparse Representation for Multiple Kernel Dimensionality Reduction

In the fields of pattern recognition and data mining, two problems need to be addressed. First, the curse of dimensionality degrades the performance of many practical data processing techniques. Second, due to the existence of noise and outliers, feature extraction on corrupted data cannot be effectively achieved. Recently, some representation based methods have produced promising results. However, these methods cannot handle the case in which nonlinear similarity exists and have failed to provide the quantized interpretability for the importance of features. In this paper, we propose a novel low rank and sparse representation method to realize dimensionality reduction and robustly extract latent low dimensional discriminative features. Specifically, we first adopt multiple kernel learning to map the original data into an embedded reproducing kernel Hilbert space (RKHS) and then kernel based similarity discriminative projection is learned to explore the within-class and between-class variability. Notably, this low dimensional feature learning strategy is definitely integrated into the low rank matrix recovery of the kernel matrix. Next, we introduce the regularization of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$l_{2,1}$ </tex-math></inline-formula> norm on error matrix to eliminate noise and on projection matrix to lead the selected features to be more compact and interpretable. The non-convex optimization problem is effectively solved by the alternating direction method of multipliers (ADMM) methods. Extensive experiments on seven benchmark datasets are conducted to demonstrate the effectiveness of our method.

Understanding Deep Learning Techniques for Recognition of Human Emotions Using Facial Expressions: A Comprehensive Survey

Emotion recognition plays a significant role in cognitive psychology research. However, measuring emotions is a challenging task. Thus, several approaches have been designed for facial expression recognition (FER). Although, the challenges increases further as the data transits from laboratory-controlled environment to <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in-the-wild</i> circumstances. Nowadays, applications are overwhelmed by a profusion of deep learning (DL) techniques in the real-world problems. DL networks have steadily led to a better understanding of low-dimensional discriminative features from high-dimensional complex face patterns for automatic FER. The modern FER systems based on deep neural networks mainly suffer from two problems: overfitting due to the inadequate availability of training data and complications unassociated with the expressions, such as occlusion, posture, illumination, and identity bias. This study aims to provide a comprehensive survey of the significant DL-based methods that have made a notable contribution to the field of FER. Different components of the methods, such as pre-processing, feature extraction, and classification of facial expressions, are described systematically. Moreover, the discussed approaches are analyzed to compare their performance along with their advantages and limitations. Further, different databases relevant to FER are also explored in this study. Essentially, the main aim of this survey is twofold. The former is to discuss the current scenario of FER approaches and the latter is to present some thoughts on the future directions of facial emotion recognition by machines: what are the obstacles and prospects for FER researchers?.

Cyber–physical cloud-oriented multi-sensory smart home framework for elderly people: An energy efficiency perspective

The emerging Cyber–Physical Systems (CPSs) provide a number of services ranging from smart homes to elderly care. With a CPS, a user’s (elderly person’s) interactions with smart home appliances are sensed, collected, and shared in physical spaces. This interaction is then delivered to cyberspaces for processing in order to monitor energy efficiency. In this paper, we propose an energy-efficient cyber–physical smart home system for monitoring the elderly that uses the potential of cloud computing and big data technologies. Given that an elderly person may need assistance with the complex activities of maintaining energy efficiency, a smart multimedia-enabled middleware assistant is proposed that enables him/her to observe different energy-efficient processes, control smart home appliances through gestures, receive notifications regarding appliance statuses, and share multimedia messages with a community of interest. For data processing, robust and low-dimensional discriminative features are extracted in a server from the data available via sensors and multimedia. Events are also automatically classified in the server to assist caregiver decision-making. To save energy, an automatic switch on/off control system using speech or gesture is developed. Keywords are detected from the speech and used as control commands. Experimental results show the encouraging potential of deploying this framework at a large scale.

Analysis and Classification of Natural Rock Textures based on New Transform-based Features

This work develops a mathematical method to extract relevant information about natural rock textures to address the problem of automatic classification. Classical methods of texture analysis cannot be directly applied in this context, since rock textures are typically characterized by both stationary patterns (a classic kind of texture) and geometric forms, which are not properly captured with conventional methods. Due to the presence of these two phenomena, a new classification approach is proposed in which each rock texture class is individually analyzed developing a specific low-dimensional discriminative feature. For this task, multi-scale transform domain representations are adopted, allowing the analysis of the images at several levels of scale and orientation. The proposed method is applied to a database of digital photographs acquired in a porphyry copper mining project, showing better performance than state-of-the-art techniques, and additionally presenting a low computational cost.

Higher order partial least squares for object tracking: A 4D-tracking method

Object tracking has a wide range of applications and great efforts have been spent to build the object appearance model using image features encoded in a vector as observations. Since a video or image sequence is intrinsically a multi-dimensional matrix or a high-order tensor, these methods cannot fully utilize the spatial-temporal correlations within the 2D image ensembles and inevitably lose a lot of useful information. In this paper, we propose a novel 4D object tracking method via the higher order partial least squares (HOPLS) which is a generalized multi-linear regression method. To do so, we first represent each training and testing example as a set of image instances of a target or background object. Then, we view object tracking as a multi-class classification problem and construct the 4D data matrix and 2D labeling matrix for HOPLS. Furthermore, we use HOPLS to adaptively learn low-dimensional discriminative feature subspace for object representation. Finally, a simple yet effective updating schema is used to update the object appearance model. Experimental results on challenging video sequences demonstrate the robustness and effectiveness of the proposed 4D tracking method.

Adaptive ranking of perceptual aesthetics

As humans, we love to rank things. Top ten lists exist for everything from movie stars to scary animals. Ambiguities (i.e. ties) naturally occur in the process of ranking when people feel they cannot distinguish two items. Human reported rankings derived from star ratings abound on recommendation websites such as Yelp and Netflix. However, those websites differ in star precision which points to the need for ranking systems that adapt to an individual user׳s preference sensitivity. In this work we propose an adaptive system that allows for ties when collecting ranking data. Using this system, we propose a framework for obtaining computer-generated rankings. We test our system and a computer-generated ranking method on the problem of evaluating facial aesthetics. Since aesthetics is a personalized and subjective issue, and it is hard to obtain large amount of aesthetics rankings from each user, we extract low-dimensional discriminative features from weakly labelled facial images and apply them to afterward learning. Extensive experimental evaluations and analysis well demonstrate the effectiveness of our work.

3D object tracking via image sets and depth-based occlusion detection

Object tracking has been typically formulated as an online learning problem, in which a target appearance is updated adaptively using the single-shot images tracked from the previous 2D frames. The traditional methods have the following limitations: (1) useful depth information is ignored; (2) each training and testing example is a single image, so it can be easily corrupted. In this paper, we propose a novel 3D object tracking method using image sets and depth-based occlusion detection, in which each training and testing example contains a set of image instances of an object and covers large variations in the object׳s appearance. To do so, we first represent each image set as its natural second-order statistic. Then, we use kernel partial least squares to adaptively learn low-dimensional discriminative feature subspace for object representation. Third, to prevent improper appearance model updating during occlusions, we exploit depth information obtained from binocular video data to detect occlusion. Finally, to alleviate the tracker drifting problem caused by model update, we exploit both the ground truth appearance information of the object labeled in the initial frames and the image observations obtained online. Extensive experiments on challenging video sequences demonstrate the robustness and effectiveness of the proposed method.

Ordinary Preserving Manifold Analysis for Human Age and Head Pose Estimation

We propose in this paper an ordinary preserving manifold analysis approach for human age and head pose estimation. While a large number of manifold learning algorithms have been proposed in the literature and some of them have been successfully applied to age/pose estimation, the ordinary characteristics of the age/pose information of samples have not been fully exploited to learn the low-dimensional discriminative features for these estimation tasks. To address this, we propose an ordinary preserving manifold analysis approach to seek a low-dimensional subspace such that the samples with similar label values (i.e., small age/pose difference) are projected to be as close as possible and those with dissimilar label values (i.e., large age/pose difference) as far as possible, simultaneously. Subsequently, we learn a multiple linear regression model to uncover the relation of these low-dimensional features and the ground-truth values of samples for age/pose estimation. Experimental results on facial age estimation, gait-based human age estimation, and head pose estimation are presented to demonstrate the efficacy of our proposed approach.

Fusion of Directional Spatial Discriminant Features for Face Recognition

Abstract Feature level fusion is one of the most important techniques, used to improve the performance of a face recognition system. This paper presents an approach of fusion of directional spatial discriminant features for face recognition. The key idea of the proposed method is to fuse the facial features lie along the horizontal, vertical and diagonal directions. So that this fused feature vector can contain more discriminant information than the individual facial feature lie along single direction. However due to fusion the size of fused feature vector is become larger which may increase complexity of the classifier. To optimize this lower dimensional discriminant features are again extracted from this large fused feature vector. In our experiment, we apply G-2DFLD method on the original images to extract the discriminant features. Then original images are converted into diagonal images and another set of discriminant features, representing the diagonal information, are extracted by using the G-2DFLD method. The original and diagonal features vectors are then fused to form a large feature vector. The dimension of this large fused feature vector is then reduced by PCA method and this resultant reduced feature vector is used for classification and recognition by Radial Basis Function-Neural Networks (RBF-NN). Experiments on the AT&T (formally known as ORL database) face database indicate the competitive performance of the proposed method, as compared to some existing subspaces-based methods. Click here and insert your abstract text.

Object Tracking via Partial Least Squares Analysis

We propose an object tracking algorithm that learns a set of appearance models for adaptive discriminative object representation. In this paper, object tracking is posed as a binary classification problem in which the correlation of object appearance and class labels from foreground and background is modeled by partial least squares (PLS) analysis, for generating a low-dimensional discriminative feature subspace. As object appearance is temporally correlated and likely to repeat over time, we learn and adapt multiple appearance models with PLS analysis for robust tracking. The proposed algorithm exploits both the ground truth appearance information of the target labeled in the first frame and the image observations obtained online, thereby alleviating the tracking drift problem caused by model update. Experiments on numerous challenging sequences and comparisons to state-of-the-art methods demonstrate favorable performance of the proposed tracking algorithm.

Extracting discriminative features for CBIR

Developing low-dimensional discriminative features is crucial for content-based image retrieval (CBIR). In this paper, we present a square symmetrical local binary pattern (SSLBP) texture descriptor, which is a compact symmetrical-invariant variation of local binary pattern (LBP), then we propose a merging 2-class linear discriminant analysis (M2CLDA) method to capture low-dimensional optimal discriminative features in the projection space. M2CLDA calculates discriminant vectors with respect to each class in the one-vs.-all classification scenario and then merges all the discriminant vectors to form a projection matrix. The dimensionality of the M2CLDA space fits in with the number of classes involved. Our experiments show that the SSLBP feature is an effective variation of LBP, and the M2CLDA approach improves the performance of image retrieval and image classification observably as compared with the existed LDA approaches and takes less computation complexity than the kernel discriminant analysis (KDA) methods.