Satellite edge computing has emerged as a promising paradigm to support global-scale low-latency services by enabling remote code execution in orbit. However, unlike terrestrial networks, satellite communication is fundamentally vulnerable to eavesdropping due to its wide broadcast range, high accessibility, and long-distance propagation, which is further exacerbated by the high altitudes allowing a vast set of potential eavesdropper positions. Despite these unique risks, most prior works have either ignored security or treated it as a uniform concern across services. In this paper, we take an initial step toward secure satellite edge computing by explicitly modeling and optimizing for security heterogeneity, i.e., diverse security demands across different services and dynamic jamming strategies. We formulate a novel optimization problem that jointly minimizes average information leakage for security-sensitive services and overall energy consumption while meeting latency constraints. To solve it, we propose a stochastic optimization-based algorithm, namelySOS, which dynamically controls code offloading, resource allocation, and jamming strategy with respect to service-specific security requirements. Extensive simulations show thatSOSoutperforms existing schemes by achieving up to a 92% reduction in average information leakage with only an 18% increase in energy consumption at the same latency compared to an existing algorithm. Moreover,SOSmaintains high jamming efficiency and leakage robustness under both colluding and non-colluding attacks. Our findings highlight a fundamental yet overlooked security-energy-latency trade-off and positionSOSas a pioneering step toward security-aware satellite edge computing.

Camera traps, unmanned observation devices, and deep learning-based image recognition systems have greatly reduced human effort in collecting and analyzing wildlife images. However, data collected via above apparatus exhibits 1) long-tailed and 2) open-ended distribution problems. To tackle the open-set long-tailed recognition problem, we propose the Temporal Flow Mask Attention Network that comprises three key building blocks: 1) an optical flow module, 2) an attention residual module, and 3) a meta-embedding classifier. We extract temporal features of sequential frames using the optical flow module and learn informative representation using attention residual blocks. Moreover, we show that applying the meta-embedding technique boosts the performance of the method in open-set long-tailed recognition. We apply this method on a Korean De-militarized Zone (DMZ) dataset. We conduct extensive experiments, and quantitative and qualitative analyses to prove that our method effectively tackles the open-set long-tailed recognition problem while being robust to unknown classes.