Transposed Convolution as Alternative Preprocessor for Brain-Computer Interface Using Electroencephalogram

Ear recognition is a complex research domain within biometrics, aiming to identify individuals using their ears in uncontrolled conditions. Despite the exceptional performance of convolutional neural networks (CNNs) in various applications, the efficacy of deep ear recognition systems is nascent. This paper proposes a two-step ear recognition approach. The initial step employs deep convolutional generative adversarial networks (DCGANs) to enhance ear images. This involves the colorization of grayscale images and the enhancement of dark shades, addressing visual imperfections. Subsequently, a feature extraction and classification technique, referred to as Mean-CAM-CNN, is introduced. This technique leverages mean-class activation maps in conjunction with CNNs. The Mean-CAM approach directs the CNN to focus specifically on relevant information, extracting and assessing only significant regions within the entire image. The process involves the implementation of a mask to selectively crop the pertinent area of the image. The cropped region is then utilized to train a CNN for discriminative classification. Extensive evaluations were conducted using two ear recognition datasets: mathematical analysis of images (MAI) and annotated web ears (AWEs). The experimental results indicate that the proposed approach shows notable improvements and competitive performance: the Rank-1 recognition rates are 100.00% and 76.25% for MAI and AWE datasets, respectively.

In recent scenario speaker recognition is used in many places for security purpose. Speaker recognition is technique which can verify or identify the person who is speaking. It is different than speech recognition system. In this paper we have discussed the feature extraction techniques like Mel Frequency Cepstral Co-efficient (MFCC), Linear Predictive Coding (LPC) etc. and classification techniques like Gaussian Mixture Model (GMM), Artificial Neural Networks (ANN) etc. that is available for speaker recognition. Also the survey is about how to get better efficiency in terms of speaker recognition rate by simply modifying the existing feature extraction and classification techniques.

This paper investigates fault diagnosis in batch processes and presents a comparative study of feature extraction and classification techniques applied to a specific biotechnological case study: the fermentation process model by Birol et al. (Comput Chem Eng 26:1553-1565, 2002), which is a benchmark for advanced batch processes monitoring, diagnosis and control. Fault diagnosis is achieved using four approaches on four different process scenarios based on the different levels of noise so as to evaluate their effects on the performance. Each approach combines a feature extraction method, either multi-way principal component analysis (MPCA) or multi-way independent component analysis (MICA), with a classification method, either artificial neural network (ANN) or support vector machines (SVM). The performance obtained by the different approaches is assessed and discussed for a set of simulated faults under different scenarios. One of the faults (a loss in mixing power) could not be detected due to the minimal effect of mixing on the simulated data. The remaining faults could be easily diagnosed and the subsequent discussion provides practical insight into the selection and use of the available techniques to specific applications. Irrespective of the classification algorithm, MPCA renders better results than MICA, hence the diagnosis performance proves to be more sensitive to the selection of the feature extraction technique.

The electrocardiogram (ECG) is a non-invasive tool used to diagnose various heart conditions. Arrhythmia is one of the primary causes of cardiac arrest. Early ECG beat classification plays a significant role in diagnosing life-threatening cardiac arrhythmias. However, the ECG signal is very small, the anti-interference potential is low, and the noise is easily influenced. Thus, clinicians face challenges in diagnosing arrhythmias. Thus, a method to automatically identify and distinguish arrhythmias from the ECG signal is invaluable. In this paper, a hybrid approach based on marine predators algorithm (MPA) and convolutional neural network (CNN) called MPA-CNN is proposed to classify the non-ectopic, ventricular ectopic, supraventricular ectopic, and fusion ECG types of arrhythmia. The proposed approach is a combination of heavy feature extraction and classification techniques; hence, outperforms other existing classification approaches. Optimal characteristics were derived directly from the raw signal to decrease the time required for and complexity of the computation. Precision levels of 99.31%, 99.76%, and 99.47% were achieved by the proposed approach on the MIT-BIH,EDB, and INCART databases, respectively.

Over the last years, Convolutional Neural Networks (CNNs) have been widely used in remote sensing applications, such as marine surveillance, traffic management or road networks detection. However, since CNNs have extremely high computation, bandwith and memory requirements, the hardware implementation of a CNN on space-grade devices like FPGAs for the on-board processing of the acquired images has brought many challenges. In fact, FPGAs show great advantages in terms of reconfigurability and performance-power ratio, whichmakes them a suitable election to deploy CNNs, although the computational capabilities of this type of hardware devices available on-board is limited, so implementations have to be carefully planned. In this paper, the authors present their work towards the implementation of an efficient CNN onto a spacegrade FPGA in order to achieve the on-board processing of very-high resolution remotely sensed images as soon as the data are provided by the sensor. All this work has been conducted within the EU-funded VIDEO project. As it will be presented in this paper, the work includes the introduction of a methodology based on the project constraints, the evaluation of different state-of-the-art CNN architectures by means of a new efficiency measurement also proposed in this work, the introduction of a new efficient CNN architecture, and finally, its optimized hardware implementation by means of high-level synthesis tools.

In recent years, machine learning algorithms have been developing rapidly, becoming increasingly powerful tools in decoding physiological and neural signals. The aim of this dissertation is to develop computational tools, and especially machine learning techniques, to identify the most effective methods for feature extraction and classification of these signals. This is particularly challenging due to the highly non-linear, non-stationery, and artifact- and noise-prone nature of these signals. Among basic human-control tasks, reaching and grasping are ubiquitous in everyday life. I investigated different linear and non-linear dimensionality reduction techniques for feature extraction and classification of electromyography (EMG) during a reach-grasp-lift task. The results highlighted the advantages of completely automated feature-learning by Laplacian Eigenmaps over manual feature engineering, especially when combined with classification, to achieve high accuracy with few training samples. The ability to decode and reduce the complexity of EMG could enable new practical applications for EMG in basic science and in the clinic. It could also help design humanoid and other robots. Beyond EMG, a key objective of my dissertation was to decode brain activity during the decision-making processes that lead to voluntary action. This was based on electroencephalography (EEG) and holds the promise for improved brain-computer interfaces (BCIs), particularly related to motor imagery (MI). We developed an end-to-end convolutional neural network with attentional mechanism together with different data augmentation techniques on two benchmark MI datasets. I also collected a new dataset, recorded using high-density EEG, and containing both MI and motor execution (ME) tasks. This enabled us to directly compare the decodability of MI and ME, investigate optimal channel configurations, and much more. In particular, this facilitates the analysis and decoding of MI on the fly—online and in in real time. Another potential use of EEG is measuring brain activity of people who are floating in body-temperature water in a sensory-deprivation-tank float pod. We compared lying in the float pod versus lying in bed (a control condition). And we found differences between the two, especially in the gamma band. More research is required to understand what these findings mean for levels of stress in the float pod.

Static and dynamic analysis are widely used in malware research. Although both approaches present considerable advantages according to their specific applications, they also present some challenges in terms of the time and resources constraints when extracting features. To alleviate those constraints, recent research works explore the use of techniques such as computer vision, in detecting malicious samples and their corresponding malware family. In this research, we aim to conduct a thorough comparison between a combination set of different feature extraction and classification techniques in order to identify the optimal approach that can be used in a real-time malware detection application portable executable (PE). Multiple Machine Learning (ML) and Deep Learning (DL) algorithms were used with image-based features and non image-based features. The results obtained show high performance when using static PE header features with K-Nearest Neighbor (KNN) and Bagged Trees algorithms for bi-level classification and Image-based Fast Fourier Transform (FFT) features with AlexNet for multi-level classification.

Isolated Speech recognition systems (ISRS) have been implemented using microprocessors, digital signal processors and FPGAs and have been reported in the literature. In this paper, the study and implementation of an ISRS using Cypress Programmable System on Chip (PSoC) is presented. For the implementation, PSoC5 containing the ARM Cortex-M3 CPU is used. Recognition performance is studied using three feature extraction techniques (Zero crossing, Zero crossing with end point detection and Zero crossing with end point detection and Cochlear filter) and minimum distance classifier. The ISRS is trained and tested using 10 utterances of three words. The performance is evaluated for utterances by both single speaker and multiple speaker. From this study, it is found that the system using the Zero crossing with end point detection and Cochlear filter gives the highest recognition accuracy of 80%and 70% respectively for the utterances by single speaker and multiple speakers. Higher recognition accuracy can be obtained using more robust feature extraction and classification techniques at the cost of increase in the computation complexity and storage space. The proposed ISRS has applications in voice operated robots, appliances and access control systems.

Brain-computer interfaces (BCI) allow for the brain to communicate with electronic devices. Concerning the BCI paradigms, motor imagery uses brain signals to decode an imagined movement. However, this is a hard task given the low signal-to-noise ratio. Usually, the main steps in BCI models are pre-processing, feature extraction, and classification. In recent years, Convolutional Neural Networks (CNNs) have been gaining relevance in several areas of science due to their feature extraction, translation invariance, and parameter sharing capa-bilities. Another, more recent way of feature extraction is using attention mechanisms, which are layers of neural networks based on human attention and have the ability to highlight important features. A variation of the attention mechanism is the Convolutional Block Attention Module, which combines the CNN structure with the attention mechanism. In this work, we propose a new model that joins the core architecture of EEGNet, a compact CNN widely used in the literature, with the Convolutional Block Attention Module and residual connections. The residual connections were introduced to lower data degradation throughout the model. The results highlight the residual connection's importance for the performance of the model. The proposed model obtained a kappa result 5.2% better than the EEGNet with a p-value less than 0.01 on BCI Competition IV dataset 2a, which is a well-known dataset for Motor Imagery. Furthermore, the proposal was better than EEGNet for most subjects and had the best-worst case.

Brain-computer interface (BCI) systems are proposed as a means of communication for locked-in patients. One common BCI paradigm is motor imagery in which the user controls a BCI by imagining movements of different body parts. It is known that imagining different body parts results in event-related desynchronization (ERD) in various frequency bands. Existing methods such as common spatial patterns (CSP) and its refinement filterbank common spatial patterns (FB-CSP) aim at finding features that are informative for classification of the motor imagery class. Our proposed method is a temporally adaptive common spatial patterns implementation of the commonly used filter-bank common spatial patterns method using convolutional neural networks; hence it is called TA-CSPNN. With this method we aim to: (1) make the feature extraction and classification end-to-end, (2) base it on the way CSP/FBCSP extracts relevant features, and finally, (3) reduce the number of trainable parameters compared to existing deep learning methods to improve generalizability in noisy data such as EEG. More importantly, we show that this reduction in parameters does not affect performance and in fact the trained network generalizes better for data from some participants. We show our results on two datasets, one publicly available from BCI Competition IV, dataset 2a and another in-house motor imagery dataset.

Recent advances in deep learning have had a methodological and practical impact on brain-computer interface (BCI) research. Among the various deep network architectures, convolutional neural networks (CNNs) have been well suited for spatio-spectral-temporal electroencephalogram (EEG) signal representation learning. Most of the existing CNN-based methods described in the literature extract features at a sequential level of abstraction with repetitive nonlinear operations and involve densely connected layers for classification. However, studies in neurophysiology have revealed that EEG signals carry information in different ranges of frequency components. To better reflect these multi-frequency properties in EEGs, we propose a novel deep multi-scale neural network that discovers feature representations in multiple frequency/time ranges and extracts relationships among electrodes, i.e., spatial representations, for subject intention/condition identification. Furthermore, by completely representing EEG signals with spatio-spectral-temporal information, the proposed method can be utilized for diverse paradigms in both active and passive BCIs, contrary to existing methods that are primarily focused on single-paradigm BCIs. To demonstrate the validity of our proposed method, we conducted experiments on various paradigms of active/passive BCI datasets. Our experimental results demonstrated that the proposed method achieved performance improvements when judged against comparable state-of-the-art methods. Additionally, we analyzed the proposed method using different techniques, such as PSD curves and relevance score inspection to validate the multi-scale EEG signal information capturing ability, activation pattern maps for investigating the learned spatial filters, and t-SNE plotting for visualizing represented features. Finally, we also demonstrated our method's application to real-world problems. Based on our experimental results and analyses, we believe that the proposed multi-scale neural network can be useful for various BCI paradigms, as a starting model or as a backbone network in any new BCI experiments.

The rapid emergence of technology and big data science opened up a significant amount of work that has been carried out in the field of feature extraction and classification techniques of heart sound using various deep learning methods. Practically, medical practitioners use the same old scientific method and practice to seek out any cardiac disorders and predict any abnormality in the human heart. Heart sound normalization, denoising, segmentation, feature extraction, and classification techniques provide a suitable way of study of phonocardiography (PCG) signal analysis which eventually reduces the cost, makes the system compact, and simultaneously, can work with extensive training data. This paper mainly indulges in two parts feature extraction and classification. The proposed deep learning study for PCG signal used online available heart disease datasets, and time domain features like average energy, power, root mean square (RMS), total harmonic distortion, and zero Crossing rates are used. Statistical features used are kurtosis and skewness. The acoustic features used are Mel-frequency cepstrum coefficients (MFCCs), mel, chroma, contrast, and tonnetz. For the classification of heart sound, the proposed modified CNN-based mobile network is used. The modified CNN-based mobile network is very effective in heart sound analysis as it requires very less computation time and storage. The proposed CNN-based modified Mobile Network model attained an accuracy of 99.04 + 0.07% on the test dataset with a sensitivity of 96.8 + 0.03 % and specificity of 97.2 + 0.09%.

A Brain Computer Interface (BCI) character speller allows human-beings to directly spell characters using eye-gazes, thereby building communication between the human brain and a computer. Convolutional Neural Networks (CNNs) have achieved state-of-the-art results on the BCI character spelling accuracy. Unfortunately, to the best of our knowledge, it has not been studied whether the CNN should be designed differently to increase the spelling accuracy when the number of sensors used to acquire EEG signals is different. This paper performs an empirical study to investigate this issue. First, we show a motivational example which motivates us for this investigation. Then, we propose a method to design CNNs according to the number of sensors used in the BCI character speller. This method automatically configures a parametric CNN we have devised according to the given number of sensors. Experimental results on six datasets show that we need to design different CNNs when different number of sensors are used for the acquisition of EEG signals. Experimental results also show that our designed sensor-aware CNNs outperform other CNNs in terms of spelling accuracy in most cases. Our CNNs can increase the spelling accuracy achieved by other CNNs with up to 34%.

In the field of Brain-Computer Interface (BCI), usage of Electroencephalogram (EEG) signals provides various applications in many fields. EEG signals investigation is necessary due to an effective way of determining neurological brain disorders. These disorders include Epileptic and autism in the brain structures which are identified by the abnormal synchronous electrical discharge of the neuronal activities. The multichannel EEG signals are recorded these abnormal electrical events which can be positioned on the scalp of the brain. The aim of this research work is to classify various patterns of EEG recording samples namely autistic, epilepsy and normal using the deep learning algorithm. In order to remove the artifacts from EEG dataset, the Independent Components Analysis (ICA) technique is used to minimization the noise which is segmented and filtered using an elliptic band-pass filter. Next, EEG signal transforms to Power Spectrum Density Energy Diagrams (PSDEDs). Therefore, the techniques of feature extraction can transform the signals into various feature vectors which it automatically extracts features from the PSDED by applying Deep Convolutional Neural Networks (DCNN) and by making use of the Fully Connected (EC) layer the signals can be classified as autistic, epilepsy and normal. The classification accuracy created by the suggested DCNN model is achieved 80% for training data set validation.

This study describes a method for classifying electrocorticograms (ECoGs) based on motor imagery (MI) on the brain–computer interface (BCI) system. This method is different from the traditional feature extraction and classification method. In this paper, the proposed method employs the deep learning algorithm for extracting features and the traditional algorithm for classification. Specifically, we mainly use the convolution neural network (CNN) to extract the features from the training data and then classify those features by combing with the gradient boosting (GB) algorithm. The comprehensive study with CNN and GB algorithms will profoundly help us to obtain more feature information from brain activities, enabling us to obtain the classification results from human body actions. The performance of the proposed framework has been evaluated on the dataset I of BCI Competition III. Furthermore, the combination of deep learning and traditional algorithms provides some ideas for future research with the BCI systems.