Quality Of Extracted Features Research Articles

Self-supervised feature learning for acoustic data analysis

Acoustic surveys play a pivotal role in fisheries management. During the surveys, acoustic signals are sent into the water and the strength of the reflection, so-called backscatter, is recorded. The collected data are typically annotated manually, a process that is both labor-intensive and time-consuming, to support acoustic target classification (ATC). The primary objective of this study is to develop an annotation-free deep learning model that extracts acoustic features and improves the representation of acoustic data. For this purpose, we adopt a self-supervised method inspired by the Self DIstillation with NO Labels (DINO) model. Extracting useful acoustic features is an intricate task due to the inherent variability and complexity in biological targets, as well as environmental and technical factors influencing sound interactions. The proposed model is trained with three sampling methods: random sampling, which ignores class imbalance present in the acoustic survey data; class-balanced sampling, which ensures equal representation of known categories; and intensity-based sampling, which selects data to capture backscatter variations. The quality of extracted features is then evaluated and compared. We show that the extracted features lead to improvement, in comparison to using the untreated data, in the discriminative power of several machine learning methods (k-nearest neighbor (kNN), linear regression, multinomial logistic regression) for ATC. The improvement was measured through higher accuracy in kNN (77.55% vs. 71.93%), Macro AUC in logistic regression (0.92 vs. 0.80), and R2 in linear regression (0.69 vs. 0.45) when comparing extracted features to the untreated data. Our findings highlight the advantage of applying emerging self-supervised techniques in fisheries acoustics. This study thus contributes to the ongoing efforts to improve the efficiency of acoustic surveys in fisheries management.

Multi-Sensor Image Feature Fusion via Subspace-Based Approach Using $\ell _{1}$-Gradient Regularization

Image fusion is a technique of combining two or more images into a single image which is more informative from an interpretation point of view. With the rapid development of different synthetic aperture radar sensing satellites capturing information from the earth by measuring energy in different portions of the electromagnetic spectrum (narrow/wide-band), complementary information about the area captured by different satellites is available (e.g. high-resolution spectral and RGB images). However, the estimation of the full-resolution image may not be necessary for inference approaches, including the pixel-based classification. Instead, it is desirable to extract the relevant information embedded in the available data to improve the inference capabilities. This work proposes a computational framework to estimate features with high-spatial-resolution and appropriate spectral content by combining information from a multi-sensor system. The considered multi-sensor setup is a hyperspectral imaging system with a complementary RGB sensor. The proposed framework first extracts spatial features from the RGB image using morphological profiles. Then, the fusion model assumes that the extracted features, and the hyperspectral measurements, lie in different subspaces matrices. In addition, this work developed a joint optimization scheme to solve the feature fusion problem by integrating the alternating direction method of multipliers with the block coordinate descent method. The alternating optimization method estimates the spatial features in the fusion model by penalizing the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\ell _{1}$</tex-math></inline-formula> -norm of the spatial gradient magnitudes. The quality of extracted features is measured in terms of supervised pixel-based classification methods. Extensive simulations show that the proposed approach outperforms other state-of-the-art methods in terms of classification accuracy.

Deep Spatial-Temporal Feature Extraction and Lightweight Feature Fusion for Tool Condition Monitoring

Tool condition monitoring (TCM) is vital to maintain the quality of workpieces during machining. Recently, data-driven methods based on multisensory data have been applied to TCM. The quality of extracted features is a key to realizing a successful data-driven TCM. However, the extracted features in the previous study are focused on the multicollinearity of multisensory data, which is incapable of identifying the informative and discriminative information in the long time period aspect. This article proposed a novel method for TCM using deep spatial-temporal feature extraction and lightweight feature fusion techniques. A key to the proposed method is the extraction of multicollinearity as spatial features (SPs), and the capture of long-range dependencies and nonlinear dynamics as temporal features (TFs), to fully characterize tool wear change using multisensory data. Then, a lightweight feature fusion method is used to fuse SPs, TFs, and statistical features for further removing redundant information employing the kernel-principal component analysis. Finally, support vector machines is used to predict the tool conditions using the fusion feature. Experiments on a milling machine and a gear hobbing machine are carried out to verify the effectiveness and generalization of the proposed method respectively.

Transfer learning for image classification using VGG19: Caltech-101 image data set.

Image classification is getting more attention in the area of computer vision. During the past few years, a lot of research has been done on image classification using classical machine learning and deep learning techniques. Presently, deep learning-based techniques have given stupendous results. The performance of a classification system depends on the quality of features extracted from an image. The better is the quality of extracted features, the more the accuracy will be. Although, numerous deep learning-based methods have shown enormous performance in image classification, still due to various challenges deep learning methods are not able to extract all the important information from the image. This results in a reduction in overall classification accuracy. The goal of the present research is to improve the image classification performance by combining the deep features extracted using popular deep convolutional neural network, VGG19, and various handcrafted feature extraction methods, i.e., SIFT, SURF, ORB, and Shi-Tomasi corner detector algorithm. Further, the extracted features from these methods are classified using various machine learning classification methods, i.e., Gaussian Naïve Bayes, Decision Tree, Random Forest, and eXtreme Gradient Boosting (XGBClassifier) classifier. The experiment is carried out on a benchmark dataset Caltech-101. The experimental results indicate that Random Forest using the combined features give 93.73% accuracy and outperforms other classifiers and methods proposed by other authors. The paper concludes that a single feature extractor whether shallow or deep is not enough to achieve satisfactory results. So, a combined approach using deep learning features and traditional handcrafted features is better for image classification.

Visual high dimensional industrial process monitoring based on deep discriminant features and t-SNE

Visual process monitoring allows operators to identify and diagnose faults intuitively and quickly. The performance of visual process monitoring depends on the quality of extracted features and the performance of the visual model to visualize these features. In this study, we propose a deep model for feature extraction. First, a stacked auto-encoder is used to obtain the feature representation of the input data. Second, the feature representation is fed into a multi-layer perceptron (MLP) with the Fisher criterion as the objective function. The outputs of the MLP are the extracted discriminant features. We combine the proposed feature extraction method with the visual model consisting of t-distributed stochastic neighbor embedding (t-SNE) and a back-propagation (BP) neural network (t-SNE-BP) for visual process monitoring. In this method, an industrial data set is reorganized into a dynamic sample set containing fault trend. Subsequently, the proposed feature extraction method is used to extract discriminant features from the dynamic sample set. Finally, the t-SNE-BP model maps these discriminant features into a two-dimensional space in which the normal and fault states are represented by different regions. Visual process monitoring is performed on this two-dimensional space. The Tennessee Eastman process is used to demonstrate the performance of the proposed feature extraction and visual process monitoring methods.

SDFL-FC: Semisupervised Deep Feature Learning With Feature Consistency for Hyperspectral Image Classification

Semisupervised deep learning methods (DLMs) can mitigate the dependence on large amounts of labeled samples using a small number of labeled samples. However, for semisupervised deep feature learning (SDFL), the quality of extracted features cannot be well ensured without a certain amount of labeled samples. To address this issue, we develop the SDFL method with feature consistency (SDFL-FC) for the hyperspectral image (HSI) classification. The SDFL-FC first adopts the convolutional neural network (CNN) to extract spectral–spatial features of HSI and then uses the fully connected layers (FCLs) to model the feature consistency. Moreover, two constraints that enforce both the feature consistency of single pixel (FCS) and feature consistency of group pixels (FCG) are introduced to obtain the representative and discriminative features. The FCS is achieved by the generative adversarial network (GAN) regularization, which can reconstruct the original data from extracted features. The FCG is based on the assumption that the features of group pixels should have similar characteristics within a superpixel, which is embedded in each FCL. The final FCL outputs the class labels, and the cross-entropy (CE) loss is calculated with the labeled samples, while the two losses of FCS and FCG are calculated with all the training samples (both labeled and unlabeled). SDFL-FC integrates the FCS, FCG, and CE loss into a unified objective function and uses a customized iterative optimization algorithm to optimize it. Experiments demonstrate that the SDFL-FC can outperform the related state-of-the-art HSI classification methods.

Emotion Recognition of Manipuri Speech using Convolution Neural Network

over the recent years much advancement are made in terms of artificial intelligence, machine learning, human-machine interaction etc. Voice interaction with the machine or giving command to it to perform a specific task is increasingly popular. Many consumer electronics are integrated with SIRI, Alexa, cortana, Google assist etc. But machines have limitation that they cannot interact with a person like a human conversational partner. It cannot recognize Human Emotion and react to them. Emotion Recognition from speech is a cutting edge research topic in the Human machines Interaction field. There is a demand to design a more rugged man-machine communication system, as machines are indispensable to our lives. Many researchers are working currently on speech emotion recognition(SER) to improve the man machines interaction. To achieve this goal, a computer should be able to recognize emotional states and react to them in the same way as we humans do. The effectiveness of the speech emotion recognition(SER) system depends on quality of extracted features and the type of classifiers used . In this paper we tried to identify four basic emotions: anger, sadness, neutral, happiness from speech. Here we used audio file of short Manipuri speech taken from movies as training and testing dataset . This paper use CNN to identify four different emotions using MFCC (Mel Frequency Cepstral Coefficient )as features extraction technique from speech.

Modeling Large-Scale Industrial Processes by Multiple Deep Belief Networks With Lower-Pressure and Higher-Precision for Status Monitoring

Typically, fault detection using deep learning is performed based on the features extracted from only one well-trained deep model. However, our results show that large-scale data is complicated and originates from different schemas, which will cause great pressure on deep neural networks, furthermore, the quality of the extracted features will be affected, and the training complexity and time will also be increased. Conversely, deep models would feel comfortable to extract features from raw data that contain less complex relationships and the quality of extracted features are higher and more representative. Hence, variables from large-scale industrial processes in this study are reasonably divided into various schemas with simple relationships by mutual information. Then, the corresponding deep belief network (DBN) models are established under a lighter pressure state to sufficiently extract the abstract and high-order information from data in each schema. Experimental analysis shows that the training efficiency, the accuracy of extracted features and the monitoring performance based on the proposed model system are all better than using only one DBN. What’s more, a comparison with those of representative and state-of-the-art methods on numerical and Tennessee Eastman processes also demonstrates the high performance of the proposal called M-DBN.

Statistical Process Control with Intelligence Based on the Deep Learning Model

Statistical process control (SPC) is an important tool of enterprise quality management. It can scientifically distinguish the abnormal fluctuations of product quality. Therefore, intelligent and efficient SPC is of great significance to the manufacturing industry, especially in the context of industry 4.0. The intelligence of SPC is embodied in the realization of histogram pattern recognition (HPR) and control chart pattern recognition (CCPR). In view of the lack of HPR research and the complexity and low efficiency of the manual feature of control chart pattern, an intelligent SPC method based on feature learning is proposed. This method uses multilayer bidirectional long short-term memory network (Bi-LSTM) to learn the best features from the raw data, and it is universal to HPR and CCPR. Firstly, the training and test data sets are generated by Monte Carlo simulation algorithm. There are seven histogram patterns (HPs) and nine control chart patterns (CCPs). Then, the network structure parameters and training parameters are optimized to obtain the best training effect. Finally, the proposed method is compared with traditional methods and other deep learning methods. The results show that the quality of extracted features by multilayer Bi-LSTM is the highest. It has obvious advantages over other methods in recognition accuracy, despite the HPR or CCPR. In addition, the abnormal patterns of data in actual production can be effectively identified.

High-Quality Wavelets Features Extraction for Handwritten Arabic Numerals Recognition

Arabic handwritten digit recognition is the science of recognition and classification of handwritten Arabic digits. It has been a subject of research for many years with rich literature available on the subject. Handwritten digits written by different people are not of the same size, thickness, style, position or orientation. Hence, many different challenges have to overcome for resolving the problem of handwritten digit recognition. The variation in the digits is due to the writing styles of different people which can differ significantly. Automatic handwritten digit recognition has wide application such as automatic processing of bank cheques, postal addresses, and tax forms. A typical handwritten digit recognition application consists of three main stages namely features extraction, features selection, and classification. One of the most important problems is feature extraction. In this paper, a novel feature extraction approach for off-line handwritten digit recognition is presented. Wavelets-based analysis of image data is carried out for feature extraction, and then classification is performed using various classifiers. To further reduce the size of training data-set, high entropy subbands are selected. To increase the recognition rate, individual subbands providing high classification accuracies are selected from the over-complete tree. The features extracted are also normalized to standardize the range of independent variables before providing them to the classifier. Classification is carried out using k-NN and SVMs. The results show that the quality of extracted features is high as almost equivalently high classification accuracies are acquired for both classifiers, i.e. k-NNs and SVMs.

Cross-modal hashing with semantic deep embedding

Cross-modal hashing has demonstrated advantages on fast retrieval tasks. It improves the quality of hash coding by exploiting semantic correlation across different modalities. In supervised cross-modal hashing, the learning of hash function replies on the quality of extracted features, for which deep learning models have been adopted to replace the traditional models based on handcraft features. All deep methods, however, have not sufficiently explored semantic correlation of modalities for the hashing process. In this paper, we introduce a novel end-to-end deep cross-modal hashing framework which integrates feature and hash-code learning into the same network. We take both between and within modalities data correlation into consideration, and propose a novel network structure and a loss function with dual semantic supervision for hash learning. This method ensures that the generated binary codes keep the semantic relationship of the original data points. Cross-modal retrieval experiments on commonly used benchmark datasets show that our method yields substantial performance improvement over several state-of-the-art hashing methods.

An investigation on the degradation of different features extracted from the compressed American English speech using narrowband and wideband codecs

Speech coding facilitates speech compression without perceptual loss that results in the elimination or deterioration of both speech and speaker specific features used for a wide range of applications like automatic speaker and speech recognition, biometric authentication, prosody evaluations etc. The present work investigates the effect of speech coding in the quality of features which include Mel Frequency Cepstral Coefficients, Gammatone Frequency Cepstral Coefficients, Power-Normalized Cepstral Coefficients, Perceptual Linear Prediction Cepstral Coefficients, Rasta-Perceptual Linear Prediction Cepstral Coefficients, Residue Cepstrum Coefficients and Linear Predictive Coding-derived cepstral coefficients extracted from codec compressed speech. The codecs selected for this study are G.711, G.729, G.722.2, Enhanced Voice Services, Mixed Excitation Linear Prediction and also three codecs based on compressive sensing frame work. The analysis also includes the variation in the quality of extracted features with various bit-rates supported by Enhanced Voice Services, G.722.2 and compressive sensing codecs. The quality analysis of extracted epochs, fundamental frequency and formants estimated from codec compressed speech was also performed here. In the case of various features extracted from the output of selected codecs, the variation introduced by Mixed Excitation Linear Prediction codec is the least due to its unique method for the representation of excitation. In the case of compressive sensing based codecs, there is a drastic improvement in the quality of extracted features with the augmentation of bit rate due to the waveform type coding used in compressive sensing based codecs. For the most popular Code Excited Linear Prediction codec based on Analysis-by-Synthesis coding paradigm, the impact of Linear Predictive Coding order in feature extraction is investigated. There is an improvement in the quality of extracted features with the order of linear prediction and the optimum performance is obtained for Linear Predictive Coding order between 20 and 30, and this varies with gender and statistical characteristics of speech. Even though the basic motive of a codec is to compress single voice source, the performance of codecs in multi speaker environment is also studied, which is the most common environment in majority of the speech processing applications. Here, the multi speaker environment with two speakers is considered and there is an augmentation in the quality of individual speeches with increase in diversity of mixtures that are passed through codecs. The perceptual quality of individual speeches extracted from the codec compressed speech is almost same for both Mixed Excitation Linear Prediction and Enhanced Voice Services codecs but regarding the preservation of features, the Mixed Excitation Linear Prediction codec has shown a superior performance over Enhanced Voice Services codec.

SAR Automatic Target Recognition Using a Roto-Translational Invariant Wavelet-Scattering Convolution Network

The algorithm of synthetic aperture radar (SAR) for automatic target recognition consists of two stages: feature extraction and classification. The quality of extracted features has significant impacts on the final classification performance. This paper presents a SAR automatic target classification method based on the wavelet-scattering convolution network. By introducing a deep scattering convolution network with complex wavelet filters over spatial and angular variables, robust feature representations can be extracted across various scales and angles without training data. Conventional dimension reduction and a support vector machine classifier are followed to complete the classification task. The proposed method is then tested on the moving and stationary target acquisition and recognition (MSTAR) benchmark data set and achieves an average accuracy of 97.63% on the classification of ten-class targets without data augmentation.

Quality-Based Learning for Web Data Classification

The types of web data vary in terms of information quantity and quality. For example, some pages contain numerous texts, whereas some others contain few texts; some web videos are in high resolution, whereas some other web videos are in low resolution. As a consequence, the quality of extracted features from different web data may also vary greatly. Existing learning algorithms on web data classification usually ignore the variations of information quality or quantity. In this paper, the information quantity and quality of web data are described by quality-related factors such as text length and image quantity, and a new learning method is proposed to train classifiers based on quality-related factors. The method divides training data into subsets according to the clustering results of quality-related factors and then trains classifiers by using a multi-task learning strategy for each subset. Experimental results indicate that the quality-related factors are useful in web data classification, and the proposed method outperforms conventional algorithms that do not consider information quantity and quality.

Improving passive radio‐frequency identification localisation precision with moving direction estimation‐based feature improvement

To achieve high localisation accuracy and precision is a fundamental requirement of passive radio-frequency identification (RFID) location systems. Current passive RFID location systems mostly deploy a dense RFID tag distribution to give initial localisation accuracy and use advanced localisation algorithms to extract the reliable features for enhancing localisation precision. High quality features are capable of significantly improving precision of passive RFID localisation systems, but it is hardly achieved in practice because of the environmental noise interference and the variety of tag behaviours. A localisation approach by using moving direction estimation to improve the quality of extracted features is proposed for enhancing localisation precision of passive RFID location systems. This approach relies on a general distribution of false-reading occurring probability function derived from experimental measurements. A feature adjustment scheme is introduced to correct these features regarding the estimation of moving direction of object. The experimental results demonstrate that the proposed localisation approach with feature improvement achieves higher precision for the state-of-the art deterministic localisation algorithm (Han et al. 2007) in passive RFID location systems.

A Drowsiness Detection Architecture using Feature Extraction Methodology

Drowsiness is considered as a significant risk factor that contributes to large number of accidents. Therefore a method is developed for counteracting its effects and hence a very accurate classification of drowsiness is needed so as to categorize the different states of drowsiness and thus the person can be alerted at definite instants. The quality of extracted features is assessed on datasets collected from various subjects irrespective of age, sex and also the start and stop of the signal monitoring time of the subjects. An optimal Fuzzy Entropy Mutual Information(FEMI) methodology is implemented into Wavelet Packet Transform(WPT) decomposition so as to extract the features and the various instants of drowsiness of the subject is obtained. The Receiver Operating Characteristics (ROC) curve shows the classification accuracy of the SVM classifier in this proposed methodology.

Theoretical analysis on feature extraction capability of class-augmented PCA

In this paper, we present a theoretical analysis on a novel supervised feature extraction method called class-augmented principal component analysis (CA-PCA), which is composed of processes for encoding the class information, augmenting the encoded information to data, and extracting features from class-augmented data by applying PCA. Through a combination of these processes, CA-PCA can extract features appropriate for classification. Our theoretical analysis aims to clarify the role of these processes and to provide an explanation on how CA-PCA can extract good features. Experimental results for various datasets are provided in order to show the validity of the proposed method for real problems. The effect of parameters on the quality of extracted features is also investigated and the rules of thumb for determining the appropriate parameters are provided.

Feature extraction and image segmentation using self-organizing networks

Feature extraction and image segmentation (FEIS) are two primary goals of almost all image-understanding systems. They are also the issues at which we look in this paper. We think of FEIS as a multilevel process of grouping and describing at each level. We emphasize the importance of grouping during this process because we believe that many features and events in real images are only perceived by combining weak evidence of several organized pixels or other low-level features. To realize FEIS based on this formulation, we must deal with such problems as how to discover grouping rules, how to develop grouping systems to integrate grouping rules, how to embed grouping processes into FEIS systems, and how to evaluate the quality of extracted features at various levels. We use self-organizing networks to develop grouping systems that take the organization of human visual perception into consideration. We demonstrate our approach by solving two concrete problems: extracting linear features in digital images and partitioning color images into regions. We present the results of experiments on real images.