Monitoring real-time air quality is essential for safeguarding public health and fostering social progress. However, the widespread deployment of air quality monitoring stations is constrained by their significant costs. To address this limitation, we introduce AirRadar, a deep neural network designed to accurately infer real-time air quality in locations lacking monitoring stations by utilizing data from existing ones. By leveraging learnable mask tokens, AirRadar reconstructs air quality features in unmonitored regions. Specifically, it operates in two stages: first capturing spatial correlations and then adjusting for distribution shifts. We validate AirRadar’s efficacy using a year-long dataset from 1,085 monitoring stations across China, demonstrating its superiority over multiple baselines, even with varying degrees of unobserved data.

Cameras hold great potential in sleep monitoring and are suitable for long-term home use. However, privacy invasion is a major concern of cameras in sleep monitoring. A defocused camera offers a promising privacy-protection solution, its feasibility for physiological signal measurement has been shown but its performance on sleep staging remains unknown. We applied four levels of post-blurring (from sharp to blurred videos) to investigate the sleep staging including wakefulness (Wake), rapid eye movement (REM), light stage (Light), and deep stage (Deep). The results showed that all four levels of blurring effectively provide identity protection, but the sleep staging performance (especially the Light) was degraded by video blurring, particularly due to the distorted motion cues and high-frequency components of heart rate variability (HRV). To this end, we introduced Cardiopulmonary Coupling (CPC) to video-based sleep staging. With CPC, the accuracy (ACC) reached 69.5% (kappa = 0.59, F1-score = 0.67) with privacy protection, while under the most severe blurring, ACC was improved from 59.8% (kappa = 0.49, F1-score = 0.56) to 66.9% (kappa = 0.54, F1-score = 0.65). The main benefit of CPC is that the photoplethysmography (PPG) derived respiration (PDR) signal is more robust than inter-beat interval (IBI) under video blurring. Overall, this study explores the opportunity of using blurred videos for privacy-protected sleep staging and demonstrates the effectiveness of CPC in mitigating the degradation of motion and HRV features caused by video blurring. Cohort diversity and multimodal integration are expected to offer deeper clinical and technical insights for real-world applications.

The classic problem of node importance estimation has been conventionally studied with homogeneous network topology analysis. To deal with practical network heterogeneity, a few recent methods employ graph neural models to automatically learn diverse sources of information. However, the major concern revolves around that their fully adaptive learning process may lead to insufficient information exploration, thereby formulating the problem as the isolated node value prediction with underperformance and less interpretability. In this work, we propose a novel learning framework namely SKES. Different from previous automatic learning designs, SKES exploits heterogeneous structural knowledge to enrich the informativeness of node representations. Then based on a sufficiently uninformative reference, SKES estimates the importance value for any input node, by quantifying its informativeness disparity against the reference. This establishes an interpretable node importance computation paradigm. Furthermore, SKES dives deep into the understanding that "nodes with similar characteristics are prone to have similar importance values" whilst guaranteeing that such informativeness disparity between any different nodes is orderly reflected by the embedding distance of their associated latent features. Extensive experiments on three widely-evaluated benchmarks demonstrate the performance superiority of SKES over several recent competing methods.

Measuring the fidelity of synthesized virtual traffic has become an important and fundamental concern for evaluating the performance of different traffic simulation techniques and applications of autonomous vehicle testing. In this work, we propose a novel method to evaluate the fidelity of any trajectory data from the perspective of anomalous trajectory detection. First, given the trajectory data to be evaluated as input, the method learns spatio-temporal traffic features and reconstructs the input trajectory through a Long Short-Term Memory (LSTM)-based autoencoder architecture. Then, the anomalous trajectories are detected by comparing the reconstructed trajectories and the input ones using the reconstruction error as the benchmark. Our method can detect eight different kinds of anomalous trajectory in terms of changes in velocity and moving direction. In order to evaluate the fidelity of the input trajectory, we design a perceptual evaluation on virtual traffic fidelity and derive a mapping from the reconstruction error to the evaluation score. We demonstrated the effectiveness and robustness of our metric through many experiments on real-world and synthetic trajectory data containing different types of motion anomalies.