To construct the micro collision avoidance intelligent unit, efficient solutions have been pursued by modeling a specific collision-sensitive neuron, the lobula giant movement detector (LGMD), identified in the locust’s visual neural system. However, the existing LGMD models cooperating with a threshold-based collision avoidance strategy are less robust in the presence of visual variability, including fluctuations in contrast and noise. Intriguingly, biological research underscores the potential of the single-peak modeling through the η function, suggesting a non-threshold collision avoidance strategy by identifying the peak response moment. Adapting the biological finding to the engineering applications, we establish a novel modeling method based on η-form perceptron, making the response curve behave like the η function. The single peak within the η-function-form response occurs before the collision, offering reliable collision warning cues. Numerical experiments substantiate the effectiveness of the proposed model against visual variability. This study extends the limits of the applications of LGMD models in diverse scenarios, and explores the potential of modeling neural computation in constructing micro-intelligent units.
The Liuqu Formation in southern Tibet, China, archives the late Palaeocene Liuqu flora (ca. 56 Ma) and represents ancient vegetation on the northern margin of the Indian Plate near the India–Asia collision zone. It therefore provides critical evidence for evaluating early biotic interactions between India and Eurasia. Here, we report on Syzygium (Myrtaceae) fossil leaves from the Liuqu flora. Detailed morphological comparisons suggest that these specimens represent a new fossil species, namely Syzygium ovatum Y. Gao et T. Su sp. nov. By integrating fossil records with deep-time ecological niche modelling, we infer that Syzygium originated on the Indian Plate before or during the Palaeocene. By the middle Eocene, Syzygium had migrated to the northeast along the Asian Pacific coastline to East Asia, consistent with an “Out-of-India” dispersal scenario. Fossil evidence demonstrates that the genus reached New Zealand and Australia by the Early Miocene, likely through a Southeast Asian dispersal corridor. During the Pliocene, the distribution of Syzygium resembled that of its modern range. Although the African fossil record of Syzygium only dates to the early Pliocene, climatic suitability and favourable paleogeographic conditions suggest an earlier arrival: 1) during the Palaeocene via the Kohistan–Ladakh Island Arc and associated island systems or 2) during the Oligocene, when more suitable areas for Syzygium linked Africa and India. Generally, our results suggest a Palaeocene origin of Syzygium on the Indian Plate instead of an early Eocene Australian origin as previously proposed.
Fluid-to-fluid scaling provides an effective approach for investigating thermo-hydraulic systems employing working fluids that are difficult to study experimentally. In this study, the scaling methodology is applied to single-phase natural circulation loops (NCLs) and evaluated for both liquid–liquid (water–molten salt) and liquid–gas (water–helium) systems. A CFD model validated against available integral experimental data for the reference water loop is used to assess the scaling performance. Global similarity is assessed through dimensionless similarity criteria, while local thermo-hydraulic fields are analyzed to identify the mechanisms responsible for similarity distortions. The results show that the scaling law preserves the dominant global thermo-hydraulic characteristics for both fluid pairings, with the deviations of the principal operating parameters from their target similarity values remaining within approximately 3.5%. Moderate distortions in the secondary-side heat-transfer condition have a limited influence on the global circulation response. The remaining local distortions arise primarily from differences in thermo-physical properties that are not explicitly constrained by the scaling criteria, leading to more pronounced temperature and velocity deviations in the liquid–gas case. These deviations are associated mainly with differences in thermal diffusion and the resulting redistribution of buoyancy across the flow cross-section. The results clarify the applicability of unified fluid-to-fluid scaling for systems with substantially different thermo-physical properties and support the development of practical experimental approaches for such systems.
Multimodal pathological diagnosis is a critical paradigm for accurate clinical decision making. Despite substantial progress, existing weakly supervised multimodal pathological diagnostic approaches integrating whole-slide images(WSIs) and structured clinical data typically rely on predefined collaboration strategies that assume uniform modality reliability, failing to account for sample-wise modality-specific noise and quality variations. Under weak supervision, the lack of fine-grained instance-level annotations prevents reliable assessment of sample-wise modality reliability, causing unreliable modalities to be indiscriminately fused and amplified, leading to unstable representation learning and unreliable multimodal decision making. To address this challenge, we propose a confidence-based multimodal diagnostic framework that models sample-wise modality reliability via prediction confidence and incorporates it throughout representation learning and multimodal decision making. Specifically, a dual-path global-local attention mechanism suppresses attention noise and stabilizes confidence estimation in pathological image representations, while large language models(LLMs)-based semantic reasoning improves the quality and consistency of structured clinical data through clinically plausible missing value imputation. At the multimodal decision stage, prediction confidence is used as an indicator of modality reliability to guide adaptive fusion, preventing unreliable modalities from dominating predictions. Evaluated on CAMELYON16, TCGA-BRCA, and TCGA-NSCLC datasets, the proposed model demonstrates robust improvements. On the TCGA-NSCLC dataset, it achieves 93.36% Accuracy and 96.40% AUC, outperforming the best unimodal baseline by 4.86% and 2.11%. Furthermore, it consistently surpasses multimodal approaches with up to 1.42% accuracy improvements, proving its superior robustness under sample-wise modality noise. GitHub repository: https://github.com/KuaLe/Pathologic_Fusion.
