Saliency-based explainable AI (XAI) methods are commonly used to explain the behaviors of AI models, despite the limited research on whether such methods can indeed enhance user understanding. Here we proposed a set of tasks to systematically and objectively evaluate user’s global understanding of object detection models at the feature, object, and image levels. We found that while presenting AI’s hits, misses, and false alarms to users could enhance feature-level and some aspects of object-level understanding, presenting saliency-based explanations could not provide any additional help and did not help direct user’s attention to relevant features. Meanwhile, presenting AI’s hits, misses, and false alarms alone did not help users distinguish AI’s hits from misses and did not enhance image-level understanding. At the image level, among the participants, assuming that AI would behave like themselves appeared to be the best strategy for predicting AI’s behavior, since any attempts to revise such assumption resulted in further deviations from AI’s actual behaviors. Thus, it is necessary to develop more effective XAI methods, particularly for object detection models. Our eye movement analyses showed that participants who used similar strategies to AI also tended to perform more similarly to AI, suggesting that we could instruct users to use their own strategy as a reference point to predict AI’s behavior accordingly. Also, participants’ eye movement consistency and attention strategy similarity to AI’s were associated with different aspects of user understanding, suggesting that eye movements could be used as non-intrusive measures to monitor user understanding for providing user-specific explanations in future XAI methods.

Automatic seizure detection holds significant importance for epilepsy diagnosis and treatment. Convolutional neural networks (CNNs) have shown immense potential in seizure detection. Though traditional CNN-based seizure detection models have achieved significant advancements, they often suffer from excessive parameters and limited interpretability, thus hindering their reliability and practical deployment on edge computing devices. Therefore, this study introduces an innovative Morlet convolutional neural network (Morlet-CNN) framework with its effectiveness demonstrated in seizure detection tasks. Unlike traditional CNNs, the convolutional kernels in the Morlet-CNN contain only two learnable parameters, allowing for a lightweight architecture. Additionally, we propose a frequency-domain-response-based kernel pruning algorithm for Morlet-CNN and implement an INT8 quantization algorithm by incorporating Kullback-Leibler (KL) divergence calibration with a Morlet lookup table (LUT). With the pruning and quantization algorithms, the model's parameter scale achieves over 90% reduction while maintaining minimal accuracy loss. Furthermore, the model exhibits enhanced interpretability from a signal processing perspective, distinguishing it from many previous CNN models. Extensive experimental validation on the Bonn and CHB-MIT datasets confirms the Morlet-CNN model's efficacy with a compact Kilobyte (KB)-level model size, making it highly suitable for real-world applications.

The Motor Imagery Brain-Computer Interfaces (MI-BCIs) have shown considerable promise for applications in neural rehabilitation. However, improving the practicality and interpretability of MI-BCIs remains a critical challenge. Unlike previous methods that focus generally on either spatial, frequency, or temporal domains with coarse-grained segmentation schemes, this study proposes a novel fine-grained spatial-frequency-time (FGSFT) framework, aiming to enhance the efficiency and reliability of MI-BCIs. Multi-channel MI EEG recordings are firstly processed through multiscale time-frequency segmentation and spatial segmentation schemes, yielding fine-grained spatial-frequency-time segments (SFTSs). The key SFTSs are then selected with a tailored wrapper-based feature selection approach. Discriminative MI EEG features are extracted using a divergence-based common spatial pattern algorithm with intra-class regularization and classified using an efficient linear support vector machine (SVM). The proposed framework was evaluated on the BCI IV IIa and SDU-MI datasets, demonstrating state-of-the-art performance in terms of information transfer rate (ITR) Meanwhile, the proposed spatial segmentation strategy can significantly improve the performance of MI-BCIs when using a larger number of electrodes. Additionally, the fine-grained Motor Imagery Time-Frequency Reaction Map (MI-TFRM) and time-frequency topographical map can be obtained with the proposed framework enabling visualization of the subject-specific dynamic neural process during motor imagery tasks, facilitating the devising of personalized MI-BCIs. The FGSFT framework significantly advances the accuracy, ITR, and interoperability of MI-BCIs, paving the way for future neuroscientific research and clinical applications in neural rehabilitation and assistive technologies.

Electroencephalography (EEG) plays a crucial role in epilepsy analysis, and epileptic seizure prediction has significant value for clinical treatment of epilepsy. Currently, prediction methods using Convolutional Neural Network (CNN) primarily focus on local features of EEG, making it challenging to simultaneously capture the spatial and temporal features from multi-channel EEGs to identify the preictal state effectively. In order to extract inherent spatial relationships among multi-channel EEGs while obtaining their temporal correlations, this study proposed an end-to-end model for the prediction of epileptic seizures by incorporating Graph Attention Network (GAT) and Temporal Convolutional Network (TCN). Low-pass filtered EEG signals were fed into the GAT module for EEG spatial feature extraction, and followed by TCN to capture temporal features, allowing the end-to-end model to acquire the spatiotemporal correlations of multi-channel EEGs. The system was evaluated on the publicly available CHB-MIT database, yielding segment-based accuracy of 98.71%, specificity of 98.35%, sensitivity of 99.07%, and F1-score of 98.71%, respectively. Event-based sensitivity of 97.03% and False Positive Rate (FPR) of 0.03/h was also achieved. Experimental results demonstrated this system can achieve superior performance for seizure prediction by leveraging the fusion of EEG spatiotemporal features without the need of feature engineering.

Explainable artificial intelligence (XAI) has been increasingly investigated to enhance the transparency of black-box artificial intelligence models, promoting better user understanding and trust. Developing an XAI that is faithful to models and plausible to users is both a necessity and a challenge. This work examines whether embedding human attention knowledge into saliency-based XAI methods for computer vision models could enhance their plausibility and faithfulness. Two novel XAI methods for object detection models, namely FullGrad-CAM and FullGrad-CAM++, were first developed to generate object-specific explanations by extending the current gradient-based XAI methods for image classification models. Using human attention as the objective plausibility measure, these methods achieve higher explanation plausibility. Interestingly, all current XAI methods when applied to object detection models generally produce saliency maps that are less faithful to the model than human attention maps from the same object detection task. Accordingly, human attention-guided XAI (HAG-XAI) was proposed to learn from human attention how to best combine explanatory information from the models to enhance explanation plausibility by using trainable activation functions and smoothing kernels to maximize the similarity between XAI saliency map and human attention map. The proposed XAI methods were evaluated on widely used BDD-100K, MS-COCO, and ImageNet datasets and compared with typical gradient-based and perturbation-based XAI methods. Results suggest that HAG-XAI enhanced explanation plausibility and user trust at the expense of faithfulness for image classification models, and it enhanced plausibility, faithfulness, and user trust simultaneously and outperformed existing state-of-the-art XAI methods for object detection models.

The face recognition of familiar and unfamiliar people is an essential part of our daily lives. However, its neural mechanism and relevant electroencephalography (EEG) features are still unclear. In this study, a new EEG-based familiar and unfamiliar faces classification method is proposed. We record the multichannel EEG with three different face-recall paradigms, and these EEG signals are temporally segmented and filtered using a well-designed filter-bank strategy. The filter-bank differential entropy is employed to extract discriminative features. Finally, the support vector machine (SVM) with Gaussian kernels serves as the robust classifier for EEG-based face recognition. In addition, the F-score is employed for feature ranking and selection, which helps to visualize the brain activation in time, frequency, and spatial domains, and contributes to revealing the neural mechanism of face recognition. With feature selection, the highest mean accuracy of 74.10% can be yielded in face-recall paradigms over ten subjects. Meanwhile, the analysis of results indicates that the EEG-based classification performance of face recognition will be significantly affected when subjects lie. The time–frequency topographical maps generated according to feature importance suggest that the delta band in the prefrontal region correlates to the face recognition task, and the brain response pattern varies from person to person. The present work demonstrates the feasibility of developing an efficient and interpretable brain–computer interface for EEG-based face recognition.

Automatic seizure detection is of great significance for epilepsy diagnosis and alleviating the massive burden caused by manual inspection of long-term EEG. At present, most seizure detection methods are highly patient-dependent and have poor generalization performance. In this study, a novel patient-independent approach is proposed to effectively detect seizure onsets. First, the multi-channel EEG recordings are preprocessed by wavelet decomposition. Then, the Convolutional Neural Network (CNN) with proper depth works as an EEG feature extractor. Next, the obtained features are fed into a Bidirectional Long Short-Term Memory (BiLSTM) network to further capture the temporal variation characteristics. Finally, aiming to reduce the false detection rate (FDR) and improve the sensitivity, the postprocessing, including smoothing and collar, is performed on the outputs of the model. During the training stage, a novel channel perturbation technique is introduced to enhance the model generalization ability. The proposed approach is comprehensively evaluated on the CHB-MIT public scalp EEG database as well as a more challenging SH-SDU scalp EEG database we collected. Segment-based average accuracies of 97.51% and 93.70%, event-based average sensitivities of 86.51% and 89.89%, and average AUC-ROC of 90.82% and 90.75% are yielded on the CHB-MIT database and SH-SDU database, respectively.

Automatic seizure detection is significant for the diagnosis of epilepsy and reducing the massive workload of reviewing continuous EEGs. In this work, a novel approach, combining Stockwell transform (S-transform) with deep Convolutional Neural Networks (CNN), is proposed to detect seizure onsets in long-term intracranial EEG recordings. Primarily, raw EEG data is filtered with wavelet decomposition. Then, S-transform is used to obtain a proper time-frequency representation of each EEG segment. After that, a 15-layer deep CNN using dropout and batch normalization serves as a robust feature extractor and classifier. Finally, smoothing and collar technique are applied to the outputs of CNN to improve the detection accuracy and reduce the false detection rate (FDR). The segment-based and event-based evaluation assessments and receiver operating characteristic (ROC) curves are employed for the performance evaluation on a public EEG database containing 21 patients. A segment-based sensitivity of 97.01% and a specificity of 98.12% are yielded. For the event-based assessment, this method achieves a sensitivity of 95.45% with an FDR of 0.36/h.

A cross-sensor calibration technique is developed and applied to improve upon the prelaunch radiance calibration and characterization for the Total Ozone Unit (TOU) onboard the FengYun-3/A satellite. The Level 3 products from the National Aeronautics and Space Administration Ozone Monitoring Instrument (OMI) onboard the Earth Observing System Aura are used as input to a radiative transfer model to predict the TOU radiances and characterize the biases for the measurements over the Pacific Ocean in low- and midlatitudes. The coefficients are derived from a regression algorithm to adjust the TOU radiances. It is shown that, after the measurement bias correction, the biases between the retrieved total column ozone products from the TOU with those from the OMI Total Ozone Mapping Spectrometer (TOMS)-Version 8 products and those from a set of ground-based station measurements are 3 % and 5% , respectively. The variations in the estimated total ozone amounts from the TOU are consistent with those derived from Solar Backscatter Ultraviolet Radiometer instruments and OMI for a period from January 2010 to February 2011.