The accurate interpretation of EEG signals is a major challenge for Brain-Computer Interfaces (BCIs) based on Motor Imagery (MI). We proposed a CSP-Enhanced EEGNet model to classify Rest, Left-hand, and Right-hand motor imagery tasks. The raw signals are filtered using a band-pass filter from 8-30 Hz and a 50 Hz notch filter, followed by segmentation of the processed data into epochs of fixed length. The Common Spatial Patterns (CSP) technique is used to isolate the discriminatory activities of the motor cortex before the EEGNet architecture is trained. The network was trained for 55 epochs, and convergence was observed to be stable. The converged network achieved a training accuracy of 0.62 and a validation accuracy of 0.59, which is evidence of optimal learning with little to no overfitting. The confusion matrix showed that the model is accurate in identifying the Rest state and demonstrates a good separation of the Left and Right MI tasks. In spite of the variability of EEG data, the model achieved macro-averaged ROC-AUC scores of 0.653, 0.613, and 0.567, which show that the model adequately discriminates between classes. These findings confirm the proposed architecture as a robust, reproducible baseline for future real-time BCI development.

Brain-computer interfaces allow direct communication between the brain and external devices. They have real potential for assistive technologies and neurorehabilitation. Among the various paradigms, imagined speech decoding aims to translate silent mental speech into actionable outputs. Electroencephalography is used for imagined speech decoding due to its non-invasive nature, high temporal resolution, portability, and affordability compared to other neuroimaging modalities. The noisy and non-stationary characteristics of electroencephalography signals pose challenges for reliable classification. This study proposes a dynamic wavelet basis selection augmentation method. For each electroencephalography epoch, it adaptively chooses the most informative wavelet basis by minimizing wavelet entropy. After basis selection, data augmentation is performed by injecting Gaussian noise into the corresponding wavelet coefficients. The approach helps models handle the noise and differences in electroencephalography signals better than fixed methods. The augmented signals are classified using a convolutional neural network with channel-wise excitation mechanisms to enhance discriminative feature learning. Primary performance evaluation is conducted in an intra-subject setting using a hold-out validation protocol with trial-level separation. Additional verification is performed using cross-validation with 3, 5, and 7 folds to assess robustness. The imagined speech electroencephalography dataset comprises 32 channels, 8 stimuli, and recordings from 10 participants. The highest classification accuracy of up to 98% is achieved for the words-vowels experimental combination, with a Cohen's kappa value of 0.95. Comparatively lower performance is observed in the full eight-class classification. The proposed approach outperforms conventional augmentation strategies and static wavelet-based approaches.

Electroencephalogram (EEG) based Brain computer interface (BCI) emerges as a transformative technology with vast applications in neuroscience and rehabilitation. Imagined speech is the mental process of thinking and formulating words without vocalizing them through articulators. EEG signal is used to study imagined speech which can empower individuals with neurological impairments to communicate their thoughts effortlessly. The main challenge in decoding imagined speech is the nonstationary nature of EEG signals. Identifying robust features and scarcity of imagined speech datasets for properly training machine learning (ML) based algorithms is also a challenging task. The main objective of this study is to propose augmentation methods which mitigate data scarcity in EEG-based BCIs by introducing variations and strengthening model robustness through EEG data augmentation. The second objective is to propose a novel architecture capable of detecting variations in EEG signals for imagined speech datasets and show remarkable results. Seven diverse augmentation techniques are discussed, and the performance of the proposed model is analyzed in terms of accuracy, f1-score and kappa. The classification results are then compared with the case in which no data augmentation is used. The proposed model has shown remarkable accuracy of 91% for long words by using gaussian noise augmentation.

Brain-computer interfaces (BCIs) provide direct communication between the brain and external devices. Using electroencephalogram (EEG) sensors, BCIs are applied in assistive technologies and neuroprosthetics. Among various BCI paradigms, imagined speech-based BCI aims to decode internal speech representations from EEG signals, enabling silent communication. Decoding imagined speech is challenging due to the nonstationarity, intersubject variability of EEG signals and low signal-to-noise ratio. The proposed method uses a multilayer perceptron (MLP) integrated with a convolutional block attention module (CBAM) to enhance feature learning by refining spatial and channel-wise attention. To further improve performance, wavelet-based augmentation enhances data diversity. Phase and coherence-based functional connectivity features capture interchannel dependencies critical for imagined speech classification. The proposed wavelet-augmented phase coherence features with MLP-CBAM (WaveCoh-MLP-CBAM) framework is evaluated on an imagined speech dataset. The WaveCoh-MLP-CBAM shows superior classification accuracy, F1-score, and Cohen's kappa compared to conventional approaches. Results highlight the importance of augmentation, functional connectivity features, and attention in improving EEG-based imagined speech decoding.

Brain-computer interfaces (BCIs) enable direct communication between the brain and external devices. It provides possibilities for neurorehabilitation and assistive technologies. This study focuses on imagined speech, a challenging yet rising domain in BCI research, to decode inner speech signals from EEG data. Imagined speech aims to translate inner speech or imagined words into actionable signals without requiring overt muscle activity. Imagined speech decoding faces significant challenges, including low signal-to-noise ratios, complex inter-subject variability, and the difficulty of isolating distinct neural patterns related to specific imagined words. Phase-based feature extraction is explored to address these obstacles. The phase- based information of EEG signals, is more robust to amplitude fluctuations and noise interference. Extensive preprocessing is applied to filter out and isolate artifacts effectively. Phase-based features are then extracted from the preprocessed signals, and a machine learning model is applied to classify imagined speech accurately. Experimental results show the effectiveness of the phase-based method, showing improved accuracy in imagined word classification over traditional approaches. Hence, phase-based analysis is a powerful tool for advancing BCI performance in imagined speech applications.

Neurological disorders like stroke, epilepsy, and dementia can severely impair motor function, limiting independence and quality of life. Traditional assistive technologies, such as prosthetic limbs or voice-controlled systems, often require invasive procedures, expensive equipment, or physical efforts. These may not be feasible for many individuals, highlighting the need for AI-driven EEG analysis for risk factor prediction, diagnosis and rehabilitation. Brain-Computer Interfaces (BCIs) powered by EEG-based movement classification offer a noninvasive, affordable, and accessible alternative, enabling individuals to control devices using their thoughts alone. This study aims to develop a robust EEG-based system capable of classifying different types of hand movements, both real and imagined, using machine learning. By improving classification accuracy and reducing noise in EEG signals, this research contributes toward making hands-free interaction with technology a reality. This study will help in empowering people with disabilities and bridging the accessibility gap in assistive technologies. Feature extraction is performed using the proposed Complex STFT with Hilbert Transform (CSTFT-H) method. Machine learning models, including CNN, SVM, DT, RF and MLP are utilized to analyze EEG signals and classify them into motor imagery and motor movement. Hyperparameter tuning and k-fold validation are applied to optimize model performance.The proposed model achieved 86% accuracy for classifying all four motor activity tasks across all subjects. This would greatly benefit individuals with physical disabilities by successfully interpreting motor-related EEG data in real-time.

Brain-computer interface (BCI) allows direct communication between the brain and external devices. In this study, the motor movements (MM) of the left and right hands, along with the imagined motor movements of both hands, are investigated. One major issue is the variability and noise present in electroencephalogram (EEG) signals. Many existing machine learning (ML) methods require large datasets for accurate classi-fication. The motivation behind this study is to enhance BCI per-formance by improving the classification of motor-related EEG signals, which is crucial for applications like neurorehabilitation. This research contributes to the development of more effective BCI systems and supports individuals with physical impairments by enabling real-time interpretation of brain activity, ultimately facilitating improved neurorehabilitation strategies. The problem statement focuses on effectively classifying these signals while minimizing the impact of noise. To tackle these challenges, four key aspects of the BCI pipeline are explored: signal acquisition, preprocessing, feature extraction, and classification using various ML techniques. A robust experimental setup was designed using PsychoPy to capture high-quality EEG data, and extensive preprocessing was performed to filter out noise. Then an exhaustive exploratory data analysis and Signal Preprocessing was performed to filter out the noises and select relevant EEG channels present over the motor cortex for further processing. Findings indicate that the ML model achieved a promising accuracy of 74.7%, successfully distinguishing between motor imagery and actual movement signals. The novelty of this work lies in its unique experimental design and dataset recording, emphasizing the innovative use of ICA for feature extraction. It also introduces a systematic preprocessing pipeline to enhance the overall analysis.

Imagined speech prediction is a challenging task with significant implications for brain-computer interfaces (BCIs) and assistive communication technologies. EEG signals are highly non-stationary and thus it is extremely difficult to find any relevant information from these electroencephalogram (EEG) signals just by seeing them in time domain. Imagined speech-based BCIs can be integrated into assistive technology devices that are specifically designed for individuals with severe physical disabilities and can also play a role in rehabilitation programs for individuals recovering from conditions that affect speech and mobility. This paper proposes a novel hybrid model that combines convolutional neural network (CNN) and recurrent neural networks (RNNs) to predict imagined speech patterns from electroencephalography (EEG) signals. The integration of CNNs allows the model to capture spatial information from multi-channel EEG data, while RNNs excel at capturing temporal dependencies inherent in sequential EEG signals. The proposed model architecture starts with a 1-D CNN layer to extract spatial features from raw EEG data. The output is then fed into RNNs to capture temporal patterns and relationships over time. It achieved competitive accuracy, averaging approximately 85.25, 87.50, 81.58 and 93.92 percent for short words, long words, vowels and short-long words respectively.

EEG involves recording electrical activity generated by the brain through electrodes placed on the scalp. Imagined speech classification has emerged as an essential area of research in brain-computer interfaces (BCIs). Despite significant advances, accurately classifying imagined speech signals remains challenging due to their complex and non-stationary nature. Existing approaches often struggle with low signal-to-noise ratios and high inter-subject variability. A proposed method named imagined speech functional connectivity graph (ISFCG) is implemented to deal with these issues. The functional connectivity graphs capture the complex relationships between brain regions during imagined speech tasks. These graphs are then used to extract features that serve as inputs to various machine-learning models. The ISFCG provides an alternative representation of imagined speech signals, focusing on brain connectivity features to enhance the analysis and classification process. Also, a convolutional neural network (CNN) is proposed to learn features from these complex graphs, leading to improved classification accuracy. Experimental results on a benchmark dataset demonstrate the effectiveness of our method.

The evolution of the brain computer interface (BCI) has led to direct communication between the human brain and external devices. Within BCI, one exciting research area is the use of imagined speech. In imagined speech data, subjects think of words or sentences in their minds without vocalization. The analysis of imagined speech is challenging due to the nonstationary nature of the electroencephalogram (EEG) signal, complexity, signal quality, and noise. Thus, the relevant features for subsequent processing can be better captured by transforming the EEG data into a spectral representation. This paper introduces a novel approach for analyzing EEG signals related to imagined speech by converting these signals into spectral form using an enhanced signal spectral visualization (ESSV) technique. A novel convolutional neural network (CNN) based architecture is also proposed to classify the resulting spectral images. This method demonstrates the powerful feature extraction capabilities of CNNs, enhancing the accuracy and robustness of imagined speech recognition. The paper compares various spectral techniques, machine learning models and the effect of different frequency ranges on imagined speech classification tasks. The results are analyzed in terms of accuracy, f1 score and kappa.

Imagined speech refers to the inner articulation of words without vocalization. Augmentation techniques in Electroencephalography (EEG)-imagined speech datasets are essential to compensate for limited data, enhance model generalization, and ensure robust performance. They mitigate overfitting and improve the effectiveness of models trained on these datasets. This work demonstrates the application of diverse data augmentation techniques in imagined speech tasks, where individuals mentally simulate speech without vocalization. The proposed augmentation methods including overlapping, time warping + overlapping, gaussian noise, fourier transform surrogate (FTSurrogate), time reverse, channel shuffle, time warping + jitter and no augmentation methods aim to enhance the robustness and generalization of machine learning models. Leveraging these techniques, several classification experiments using a variety of supervised learning algorithms to distinguish imagined speech patterns are conducted. This work highlights the importance of data augmentation and machine learning (ML) in imagined speech tasks. This work has applications in cognitive neuroscience and assistive rehabilitation technologies. Our multi-layer perceptron (MLP) result has shown promising outcomes. The results are analyzed in terms of accuracy, f1 score and kappa.

Imagined speech decoding aims to decipher the speech intentions of individuals by analyzing their brain activity patterns captured through EEG recordings. This emerging technology holds great potential for revolutionizing communication for individuals with speech impairments or those unable to vocalize their thoughts. This paper presents a comprehensive exploration of imagined speech decoding using EEG signals. The study is based on the KARA ONE database, a rich collection of EEG recordings obtained from participants engaging in imagined speech tasks. This paper analyses the effect of variation in electrode location and frequency on Classification Accuracy.

Data outsourcing to the cloud is gaining popularity day by day in cloud computing. It requires a dynamic architecture that avoids loopholes and stops intruders from entering the cloud environment. To retrieve the enciphered documents from the cloud, the latest technique that is enrolled effectively is multi-keyword fuzzy search. The proposed research work uses this technique along with the n-gram corpus algorithm that helps to achieve relevancy score and sub-linear search time. This algorithmic combination is able to predict the privacy ranking measure of the retrieved data. To validate the opposed approach, simulation analysis has been performed to preserve user-oriented privacy in a cloud environment. The ranking precision of the proposed method is compared with (Fu et al. in IEEE Trans Inf Forensics Secur 12(8):1874–1884. https://doi.org/10.1109/TIFS.2017.2692728 , 2017) which uses a stemming algorithm and a matching algorithm based on the inner product. The results express the efficiency of the proposed approach by showing a relative improvement in security precision up to 55%.

The Latest emergency issues are impacting the world as a result of the new 2019 corona virus rise and pandemic. It has spread to the worldwide and afflicted people to Covid-19. The virus transmitting has regarded as being transmitted by individuals who find it a convenient disease explosion. While coughing and sneezing the infection spreads from the infectious droplets. Although in the air, these droplets will still survive and transfer the virus on to humans. In worldwide, robots have been used to alleviate the proliferation of new corona virus infections, COVID-19 with food preservation, food supplies, sanitation tasks, spraying disinfectant, temperature monitoring, hand sanitizers distributing, work on sensitizing, etc. For fast strategizing, that are considered hazardous for human beings. This paper discuss the difficulties and opportunities associated with using humanoid robots to minimize the risk of spread of COVID-19 in.public healthcare. The primary application of humanoid robots is the minimization of individual interaction in public places, and the provision of containment to hygiene, disinfectants and helping. The following discussion aims to underline the value of humanoid robot's purposes in specific and to link their use as the COVID-19 perspectives. Throughout the testing, review and diagnosis of a vulnerability and for subset of events, artificial intelligence plays a crucial role. It may be used during potential for the forecast of events but also to record the number of alternative cases, restored instances and deaths. Technology based on artificial intelligence is being used to provide outstanding services such as the detection and substitution of drugs for the care of employees by robotics for the provision of prescriptions and nutrition in clinics. It also disinfects the substances in response to the spread of Covid-19.

Identification and classification of plant diseases is a major field of study, as most people in India depend on agriculture for their main source of income and food. This is one of the reasons why the identification of plant diseases plays a significant role in agriculture. It is useful to detect plant disease through any automated technique as it eliminates a significant amount of monitoring work in large crop farms and detects the symptoms of diseases at a very early stage, i.e. when they appear on plant leaves. The framework mainly involves different concepts related to image processing, such as image acquisition, image pre-processing, and extraction of features, database formation, and artificial neural network classification. This paper covers research on various methodologies for the use of neural networks to detect plant leaf diseases.

The goal to promote human limits is for Artificial Intelligence (AI). It takes a posture on public administrations, represents the increasing availability of regaining clinical data and the rapid creation of intelligent strategies. The need to stress the need to use AI in the fight against the COVID-19 crisis. The paper outlines the main role played by Ai technologies in this unprecedented war and introduces a survey of AI methods used for multiple purposes in the fight against the outbreak of COVID-19. This paper also explains how the body temperature and coughing of the incoming person are assessed and whether the incoming person has not a protective facial mask. Should either of the above tests disqualify the participant, an alarming device invokes the local officials;the entrant may otherwise enter the premises after his/her hand has been sanitized. © 2022 Author(s).