A Moon-based sensor offers a unique view for continuous Earth observation. The 2-D turntable’s pointing performance is a critical factor influencing geolocation accuracy. The vast distance between the Earth and the Moon amplifies minor pointing errors of the turntable into significant geolocation inaccuracy. By establishing a geometric model, an analytic expression of Earth’s trajectory from the Moon-based view is derived. Three critical issues are discussed: 1) the Earth’s 18.6-year trajectory forms a$16^{\circ } \times 14^{\circ }$envelope, which determines the observation range for the sensor. The rotation angle and position of the envelope vary at different lunar locations, while its size and shape remain consistent; 2) geolocation errors caused by temporal interval vary periodically with a half-sidereal month cycle and can be compensated by calculating Earth’s velocity, while errors due to the step angle show irregular oscillations. Without calibration, both parameters can introduce geolocation errors on the scale of hundreds of kilometers. Reducing both parameters can significantly improve geolocation accuracy; and 3) even with optimization of both parameters, the geolocation accuracy cannot be reduced to within a single pixel. To achieve geolocation accuracy within design requirements, it is necessary to not only optimize these two factors but also adopt additional measures to improve precision. All these insights will inform the parameter optimization and design of the Moon-based sensor for future applications.

In many practical binary classification applications, such as financial fraud detection or medical diagnosis, it is crucial to optimize a model's performance on high-confidence samples whose scores are higher than a specific threshold, which is calculated by a given false positive rate according to practical requirements. However, the proportion of high-confidence samples is typically extremely small, especially in long-tailed datasets, which can lead to poor recall results and an alignment bias between realistic goals and loss. To address this challenge, we propose a novel loss reweighting framework called Momentum Threshold-Oriented Loss (MTOL) for binary classification tasks and propose two instantiated losses of it. Given a limited FPR range, MTOL aims to improve the recall of binary classification models at that FPR range by incorporating a batch memory queue and momentum estimation mechanisms. The MTOL adaptively estimates thresholds of FPR during the model training iterations and up-weights the loss of samples in the threshold range, with little consumption of storage and computation. Our experimental results on various datasets, including CIFAR-10, CIFAR-100, Tiny-ImageNet, demonstrate the significant effect of MTOL in improving the recall at low FPR especially in class imbalance settings. These results suggest that MTOL is a promising approach in scenarios where the model's performance in the low FPR range is of utmost importance.

The grand challenge to meet the increasing demands for food by a rapidly growing global population requires protecting crops from pests. Natural active substances play a significant role in the sustainable pests and pathogenic microbes management. In recent years, natural products- (NPs), antimicrobial peptides- (AMPs), medicinal plant- and plant essential oils (EOs)-related online resources have greatly facilitated the development of pests and pathogenic microbes control agents in an efficient and economical manner. However, a comprehensive comparison, analysis and summary of these existing web resources are still lacking. Here, we surveyed these databases of NPs, AMPs, medicinal plants and plant EOs with insecticidal, antibacterial, antiviral and antifungal activity, and we compared their functionality, data volume, data sources and applicability. We comprehensively discussed the limitation of these web resources. This study provides a toolbox for bench scientists working in the pesticide, botany, biomedical and pharmaceutical engineering fields. The aim of the review is to hope that these web resources will facilitate the discovery and development of potential active ingredients of pests and pathogenic microbes control agents.