Electroencephalography (EEG) is a non-invasive technique to record bioelectrical signals. Integrating supervised deep learning techniques with EEG signals has recently facilitated automatic analysis across diverse EEG-based tasks. However, the label issues of EEG signals have constrained the development of EEG-based deep models. Obtaining EEG annotations is difficult and requires domain experts to guide collection and labeling, and the variability of EEG signals among different subjects causes significant label shifts. To solve the above challenges, self-supervised learning (SSL) has been proposed to extract representations from unlabeled samples through well-designed pretext tasks. This article concentrates on integrating SSL frameworks with temporal EEG signals to achieve efficient representations and proposes a systematic survey of the SSL for EEG signals. In this article, (1) We introduce the concept and theory of self-supervised learning and typical SSL frameworks. (2) We provide a comprehensive survey of SSL for EEG analysis, including taxonomy, methodology, and technical details of the existing EEG-based SSL frameworks, and discuss the differences between these methods. (3) We investigate the adaptation of the SSL approach to various downstream tasks, including the task description and related benchmark datasets, and further explore its application in large-scale pre-trained foundation models for EEG signals. (4) Finally, we discuss the potential directions for future SSL-EEG research.

Federated Foundation Models (FedFMs) represent a distributed learning paradigm that fuses general competences of foundation models as well as privacy-preserving capabilities of federated learning. This combination allows the large foundation models and the small local domain models at the remote clients to learn from each other in a teacher-student learning setting. This paper provides a comprehensive summary of the ten challenging problems inherent in FedFMs, encompassing foundational theory, utilization of private data, continual learning, unlearning, Non-IID and graph data, bidirectional knowledge transfer, incentive mechanism design, game mechanism design, model watermarking, and efficiency. The ten challenging problems manifest in five pivotal aspects: “Foundational Theory,” which aims to establish a coherent and unifying theoretical framework for FedFMs. “Data,” addressing the difficulties in leveraging domain-specific knowledge from private data while maintaining privacy; “Heterogeneity,” examining variations in data, model, and computational resources across clients; “Security and Privacy,” focusing on defenses against malicious attacks and model theft; and “Efficiency,” highlighting the need for improvements in training, communication, and parameter efficiency. For each problem, we offer a clear mathematical definition on the objective function, analyze existing methods, and discuss the key challenges and potential solutions. This in-depth exploration aims to advance the theoretical foundations of FedFMs, guide practical implementations, and inspire future research to overcome these obstacles, thereby enabling the robust, efficient, and privacy-preserving FedFMs in various real-world applications.

Federated learning (FL) enables multiple clients to collaboratively train a global model while keeping local data decentralized. Data heterogeneity (non-IID) across clients has imposed significant challenges to FL, which makes local models re-optimize towards their own local optima and forget the global knowledge, resulting in performance degradation and convergence slowdown. Many existing works have attempted to address the non-IID issue by adding an extra global-model-based regularizing item to the local training but without an adaption scheme, which is not efficient enough to achieve high performance with deep learning models. In this paper, we propose a Selective Self-Distillation method for Federated learning (FedSSD), which imposes adaptive constraints on the local updates by self-distilling the global model’s knowledge and selectively weighting it by evaluating the credibility at both the class and sample level. The convergence guarantee of FedSSD is theoretically analyzed and extensive experiments are conducted on three public benchmark datasets, which demonstrates that FedSSD achieves better generalization and robustness in fewer communication rounds, compared with other state-of-the-art FL methods.

In inertial-based pedestrian navigation, anchors can effectively compensate the positioning errors originating from deviations of Inertial Measurement Units (IMUs), by putting constraints on pedestrians’ motions. However, these anchors often need to be deployed beforehand, which can greatly increase system complexity, rendering it unsuitable for emergency response missions. In this paper, we propose an anchor-based pedestrian navigation approach without any additional sensors. The anchors are defined as the intersection points of perpendicular corridors and are considered characteristics of building structures. In contrast to these real anchors, virtual anchors are extracted from the pedestrian’s trajectory and are considered as observations of real anchors, which can accordingly be regarded as inferred building structure characteristics. Then a Rao-Blackwellized particle filter (RBPF) is used to solve the joint estimation of positions (trajectory) and maps (anchors) problem. Compared with other building structure-based methods, our method has two advantages. The assumption on building structure is minimum and valid in most cases. Even if the assumption does not stand, the method will not lead to positioning failure. Several real-scenario experiments are conducted to validate the effectiveness and robustness of the proposed method.

Deep learning-based methods demonstrate promising results in continuous non-invasive blood pressure measurement, whereas those models trained on large public datasets suffer from severe performance degradation in predicting from real-world user data collected in home settings. Transfer learning has been recently introduced to personalize the pre-trained model with unseen users' data to solve the problem. However, the existing methods based on network fine-tuning for model personalization require a large amount of labeled data, lacking practicality due to labeling using a cuff-based blood pressure monitor is extremely tedious and laborious for home users. In this paper, we propose a novel few-shot transfer learning approach named FewShotBP, which addresses the above-mentioned challenges by introducing a personalization adapter at the personalization stage (i.e., the transfer learning stage), and a multi-modal spectro-temporal neural network at the pre-train stage, to bridge the gap between data-hungry models and limited labeled data in realistic scenarios. To evaluate the approach's significance, we conducted experiments using both a publicly available dataset and a real-world user experiment. The results demonstrated that the proposed approach achieves similar accuracy of blood pressure prediction with 10× less data for personalization compared with the state-of-the-art method in the public dataset and achieves a mean absolute error of 6.68 mmHg (systolic blood pressure) and 3.91 mmHg (diastolic blood pressure) with only 10 personal data samples in the real-world user experiment.

As most people suffer from the sub-health problem or chronic diseases, research on sensing and managing personal health is growing rapidly. In this paper, a user-friendly smart watch, which can sense personal physiological data and physical activity is presented. Based on the device, a smartphone software is developed, this hardware and software system can gather data and monitor personal health from the aspects of physiological conditions and physical conditions. Compared with other systems, it has the advantage of providing raw sensing data and unobtrusively continuous sensing, which can bring new opportunity to pervasive personal health monitoring, and promote wearable computing research in health monitoring and disease prevention.