Researchers from Sun Yat-sen University’s Shenzhen Campus, led by WenYuan Yang and Gege Jianga, have developed a decentralized federated learning framework, DFUN-KDF, to enhance UAV network efficiency. By leveraging federated knowledge distillation, it reduces data transmission by up to 99% while addressing model heterogeneity. A robust filtering mechanism ensures stability by eliminating faulty or malicious data. DFUN-KDF outperforms traditional methods in communication energy efficiency, adaptability, and resilience to node failures and attacks. This scalable solution offers significant potential for large-scale UAV deployments in urban management and logistics.

Accurate and efficient traffic speed prediction is crucial for improving road safety and efficiency. With the emerging deep learning and extensive traffic data, data-driven methods are widely adopted to achieve this task with increasingly complicated structures and progressively deeper layers of neural networks. Despite the design of the models, they aim to optimize the overall average performance without discriminating against different traffic states. However, the fact is that predicting the traffic speed under congestion is normally more important than the one under free flow since the downstream tasks, such as traffic control and optimization, are more interested in congestion rather than free flow. Unfortunately, most of the state-of-the-art (SOTA) models do not differentiate the traffic states during training and evaluation. To this end, we first comprehensively study the performance of the SOTA models under different speed regimes to illustrate the low accuracy of low-speed prediction. We further propose and design a novel Congestion-Aware Sparse Attention transformer (CASAformer) to enhance the prediction performance under low-speed traffic conditions. Specifically, the CASA layer emphasizes the congestion data and reduces the impact of free-flow data. Moreover, we adopt a new congestion adaptive loss function for training to make the model learn more from the congestion data. Extensive experiments on real-world datasets show that our CASAformer outperforms the SOTA models for predicting speed under 40 mph in all prediction horizons.

Can metal-based nanoparticles generated by lasers help build smarter, more immersive electronics? In the latest issue of International Journal of Extreme Manufacturing, Jun-Gyu Choi and collaborators from Ajou University and Samsung Electronics present how laser ablation in liquids enables scalable, surfactant-free nanoparticle synthesis tailored for artificial sensory and neuromorphic devices. Their work marks a breakthrough in bridging material science and intelligent electronics, paving the way for high-performance, flexible, and human-like interfaces in the next wave of extended reality technologies.

Gaining insights into the complex pathways and key cell populations involved in immune dysregulation can aid the development of therapeutic approaches to treat polytrauma, which is associated with poor patient outcomes. In a new study, researchers from the USA have utilized advanced genetic analysis tools and techniques to reveal the cellular and molecular processes involved in polytrauma-induced immune dysregulation. Their findings advance our current knowledge on polytrauma and indicate actionable targets to treat immune dysregulation.

Increasing numerical studies showed that the simplest Hubbard model on the square lattice with strong repulsion may not exhibit high-temperature superconductivity (SC). It is desired to look for other possible microscopic mechanism beyond the simplest Hubbard model to realize d-wave high-temperature SC. This study proposed that the interplay between the Su-Schrieffer-Heeger electron–phonon coupling (EPC) and the Hubbard repulsion can induce robust d-wave high-temperature SC. Using state-of-the-art density-matrix renormalization group simulations, the researchers shows that d-wave SC emerges in the Su-Schrieffer-Heeger-Hubbard model with strong Hubbard interaction and moderate EPC, paving a possible new route in understanding and looking for high-temperature SC in quantum materials.

In the paper published on Science of Traditional Chinese Medicine, the authors outline the bacteriostatic activity and mechanism of minerals containing rubidium (MCR). According to the findings, MCR inhibited Staphylococcus aureus, Listeria monocytogenes, and Escherichia coli with minimum inhibitory concentrations (MICs) of 11.95, 2.60, and 2.60 mg/mL, respectively. The inhibitory activity of MCR was insignificant against Bacillus subtilis, Salmonella typhimurium, and Helicobacter pylori at 3.25 mg/mL. Mechanistic assessments showed that MCR affected bacterial conductivity, protein and nucleic acid levels, reducing sugar content, respiratory chain dehydrogenase activity, bacterial lipid peroxidation, intracellular adenosine triphosphate, and extracellular alkaline phosphatase.

Attention detection using electroencephalogram (EEG) signals has become a popular topic. However, there seems to be a notable gap in the literature regarding comprehensive and systematic reviews of machine learning methods for attention detection using EEG signals. Therefore, this survey outlines recent advances in EEG-based attention detection within the past five years, with a primary focus on auditory attention detection (AAD) and attention level classification. First, researchers provide a brief overview of commonly used paradigms, preprocessing techniques, and artifact-handling methods, as well as listing accessible datasets used in these studies. Next, researchers summarize the machine learning methods for classification in this field and divide them into two categories: traditional machine learning methods and deep learning methods. Researchers also analyze the most frequently used methods and discuss the factors influencing each technique’s performance and applicability. Finally, researchers discuss the existing challenges and future trends in this field.

A recent review in journal Earth and Planetary Physics highlights that China's Tianwen-2 mission, launched on May 29, 2025, will carry a penetrating radar to directly probe the internal structures of the near-Earth asteroid 2016 HO₃ (Kamo'oalewa) and the active asteroid 311P/PANSTARRS. This investigation is expected to provide crucial data for unveiling the internal characteristics of asteroids and comets, thereby offering new insights into the early evolution of the solar system.

In a paper published in Earth and Planetary Physics, researchers propose a semi-empirical model combining Burton's empirical Dst formula with global magnetohydrodynamic (MHD) simulations to predict geomagnetic storm intensity. The hybrid approach demonstrates higher accuracy than pure empirical models when tested against moderate-to-intense storm events, while maintaining computational efficiency for operational space weather forecasting. This advancement enables more reliable Dst index estimation within global magnetosphere simulations.