Hong Kong University of Science and Technology
Hong Kong University of Science and Technology Institutional RepositoryNot a member yet
162821 research outputs found
Sort by
Lascoux-type resolutions, derived categories, and flips
No full textThis paper introduces Lascoux-type complexes that extend the Lascoux complexes for resolving generic determinantal ideals. These Lascoux-type complexes naturally arise when analyzing the correspondences between two different types of resolutions of singularities of determinantal varieties. We also discuss the applications of these resolutions in various geometric contexts, including blowups, standard flips, virtual flips, and projectivizations.</p
Development of shear-thickening-gel applied carbon fiber reinforced polymer (SACFRP) with enhanced low-velocity impact resistance
No full textConventional carbon fiber-reinforced polymers (CFRPs) are highly susceptible to low-velocity impact (LVI) from sharp objects due to their inherent brittleness. To address this critical limitation, an innovative shear thickening gel (STG) was incorporated into CFRP through a bespoke fabrication process, resulting in the STG-applied CFRP (SACFRP). LVI tests revealed that specific impact strength of the SACFRP increased significantly by 267 % compared to the reference CFRP fabricated with the same carbon fibers and epoxy resin but without STG. Moreover, the SACFRP achieved the specific impact strength of 202 J m/kg, substantially exceeding that of other representative carbon or glass fiber-reinforced polymers. Damage analysis and Timoshenko's theoretical study highlighted distinct failure mechanisms between the SACFRP that exhibited thin-plate elastic flexure and the CFRP that experienced brittle impact failure under LVI. Additionally, ultrasonic C-scan results demonstrated enlarged effective impact-resistant area in the SACFRP due to the viscoelasticity and shear-thickening behavior of the integrated STG, facilitating energy dissipation and reducing brittleness of the composite. In summary, this work presents the manufacturing method of an innovative SACFRP composite and demonstrates its outstanding impact resistance, marking the significant advancement in development of high-performance composites.</p
Nutrient availability controls phytoplankton populations and their nutritional strategy in the eastern Indian Ocean
No full textThe pico- and nanophytoplankton communities in the eastern Indian Ocean during the fall–winter inter-monsoon season were analyzed using flow cytometry to clarify the environmental factors that control the horizontal and vertical distributions of phytoplankton. The average Synechococcus abundance within the surface mixed layer showed a significant positive correlation with the temperature and nitrate + nitrite (N + N) concentration. Similarly, the cell concentration of eukaryotic phytoplankton in the surface mixed layer was correlated with temperature but did not decrease with decreasing N + N availability. Instead, the proportion of potentially phagotrophic eukaryotic phytoplankton, assessed using a fluorescent probe, increased with decreasing N + N concentrations in the surface mixed layer. This suggested that nitrogen uptake from particles can compensate for the decrease in inorganic nitrogen nutrients in the mixed layer, which may help eukaryotic phytoplankton maintain their biomass in oligotrophic areas. Phagotrophy by eukaryotic phytoplankton in this area may facilitate their growth, with photosynthesis driven by high irradiance within the surface mixed layer, which is depleted of nitrogen. Inter-provincial variations in cell concentrations at the subsurface peak were smaller than those within the surface mixed layer. The cell concentration of Synechococcus at the peak was positively correlated with temperature. By contrast, the peak cell concentration of eukaryotes was positively correlated with light intensity at that depth, suggesting a potential light limitation. The lower potential phagotrophy in eukaryotic phytoplankton with depth suggested that they do not use phagotrophy to compensate for diminished photosynthetic carbon acquisition.</p
Native postsynaptic density is a functional condensate formed via phase separation
No full textPhase separation is emerging as a prime mechanism in organizing dynamic subsynaptic compartments. However, studying phase separation in synapses of living neurons is challenging due to the small size of synapses. In this study, we leverage native postsynaptic densities (PSDs) purified from the mouse brain to investigate their organization. Unlike reconstituted PSDs, which form liquid-like droplets, native PSDs exhibit a gel-like morphology with defined molecular composition. Despite their morphological rigidity, native PSDs retain full molecular plasticity, manifested by selectively recruiting or excluding synaptic proteins and undergoing Ca2+-dependent structural reorganization. Notably, CaMKII in purified PSDs can be rapidly activated by Ca2+, leading to sustained phosphorylation of GluA1 and other PSD proteins. Actin polymerization further enlarges PSD clusters, mirroring structural changes during synaptic potentiation. Thus, native PSDs are functional condensates formed via phase separation. The purified PSDs also serve as an easily accessible platform for studying dynamic regulation of synapses in test tubes.<br/
Orientation-aware detection system for real-time monitoring of cracks in steel structures
No full textCrack detection plays a crucial role in steel structure health monitoring. However, conventional methods primarily rely on horizontal bounding box (HBB) detection to locate cracks, which lack orientation information and are susceptible to noise and false positives, thus hindering accurate and real-time performance. To address these limitations, this study proposes an orientation-aware detection system designed to more accurately assess the condition of steel cracks. Specifically, a parallel adaptive perceptual (PAP) attention module, an information interaction perception (I2P) head, and an orientation-shape guided (OSG) loss function are designed to enhance the performance of steel crack detection. Extensive experiments on both a custom-built steel structure crack dataset and public benchmarks demonstrate that our framework achieves a significant mAP improvement of 4.3 %-10.0 % compared to Yolov11s-obb and Yolov12s-obb. Furthermore, our model exhibits reduced computational cost relative to the baseline while achieving state-of-the-art (SOTA) performance.</p
Diffusion-Based Virtual Staining from Polarimetric Mueller Matrix Imaging
No full textPolarization, as a new optical imaging tool, has been explored to assist in the diagnosis of pathology. Moreover, converting the polarimetric Mueller Matrix (MM) to standardized stained images becomes a promising approach to help pathologists interpret the results. However, existing methods for polarization-based virtual staining are still in the early stage, and the diffusion-based model, which has shown great potential in enhancing the fidelity of the generated images, has not been studied yet. In this paper, a Regulated Bridge Diffusion Model (RBDM) for polarization-based virtual staining is proposed. RBDM utilizes the bidirectional bridge diffusion process to learn the mapping from polarization images to other modalities such as H&E and fluorescence. And to demonstrate the effectiveness of our model, we conduct the experiment on our manually collected dataset, which consists of 18,000 paired polarization, fluorescence and H&E images, due to the unavailability of the public dataset. The experiment results show that our model greatly outperforms other benchmark methods. Our data and code are available at https://github.com/xiaoyu-z/RBDM/</p
When new shamans enter the stage: ecstatic healing and the neutralisation of messianism among Akha in northwestern Laos
No full textAnaerobic co-digestion as a strategy for treating coffee Pulp: Insights into process performance
No full textCoffee pulp (CP) is an energy-rich agro-industrial waste, but its treatment in anaerobic digestion is prone to digester instability. This study evaluated the performance and microbial dynamics of mono-digestion and co-digestion systems treating CP and cattle manure (CM) at increasing organic loading rates (OLRs). At 1.0 g volatile solids (VS) L−1 d−1 , CP mono-digestion produced significantly higher methane yields than both CM mono-digestion and CP:CM co-digestion (4:1, VS basis), confirming its high energy potential under stable digester conditions. However, the methane yield substantially decreased in CP mono-digestion and CP:CM co-digestion systems at an OLR of 1.5 g VS L−1 d−1 . Extending the retention time to 30 days yielded similar methane yields and methanogenic community profiles for both CP mono-digestion and CP:CM co-digestion. Specifically, the relative abundance of acetoclastic methanogens declined while hydrogenotrophic methanogens became more prevalent. This pattern suggests that feedstock-related inhibition favored hydrogenotrophic methanogenic pathways over acetoclastic methanogenesis. These findings underscore CP’s high energy potential but also its sensitivity to elevated OLRs, emphasizing the need for strategies to counteract inhibitors.</p
Do Tropical Cyclones Have a Steady Translation Under a Uniform Steering Flow?
No full textChanges in tropical cyclone (TC) movement are commonly attributed to those in the steering flow, beta effect, or topographic influences. However, a series of idealized simulations suggest that significant track deflections can still occur even under a steady steering flow on an f plane. TCs embedded in easterly flows of varying strength systematically deflect southward from the expected westward track when radiative effects are included. The resulting track deflection reaches approximately 200 km in some experiments over a 144-hr period, comparable to typical 72- to 96-hr forecast errors in global numerical weather prediction model. A potential vorticity tendency analysis reveals that the deflection primarily results from the diabatic heating and horizontal advection terms, each linked to asymmetries in the convection and wind fields, respectively. These asymmetries are initially triggered by vortex–flow interactions and further enhanced by radiative diurnal cycles. Our findings highlight the role of internal vortex asymmetries in modulating TC motion.</p
A generative spike prediction model using behavioral reinforcement for re- establishing neural functional connectivity
No full textPrediction models that generate neuronal spikes from upstream neural activities offer a promising way to re-establish neural functional connectivity. Traditional methods train these models by supervised learning, which requires downstream recordings as ground truth. However, functional downstream activity cannot be recorded when neurological disorders exist. Here we introduce a reinforcement learning (RL)-based point process framework to generate spike trains that directly maximize behavior-level rewards, thus bypassing downstream recordings. This yields a generative spike model that directly transforms upstream activity into spike patterns modulated to desired behavior. We show that these RL-based generative models produce movement-modulated spike patterns akin to downstream recordings from healthy subjects, providing a biomimetic spike encoding framework. This RL framework outperforms existing methods and demonstrates a strong adaptation capability across different decoder settings, highlighting its potential for neural prostheses in restoring transregional communication with biomimetic cortical stimulation