Search Results (21)

Understanding how endangered carnivores partition spatiotemporal distribution in human-dominated landscapes is pivotal for mitigating biodiversity loss in climate-sensitive boreal ecosystems. Here, we used kernel density data derived from a 16-month camera-trap survey (140 UVL7 cameras), cold single-season (November–April) occupancy models, and MaxEnt 3.4.4 to identify the effects of biotic interactions, anthropogenic disturbance, and environmental factors on the spatiotemporal distribution of the wolverine (Gulo gulo) in Beijicun National Nature Reserve, Heilongjiang Province, China. We found that wolverines exhibited crepuscular activity patterns using night-time relative abundance index (NRAI) = 50.29% with bimodal peaks (05:00–07:00, 13:00–15:00), with dawn activity predominant during the warm season (05:00–06:00) and a bimodal activity pattern in the cold season (08:00–09:00, 14:00–15:00). Temporal overlap with prey (overlap coefficient Δ = 0.84) and competitors (Δ = 0.70) was high, but overlap with human-dominated temporal patterns was low (Δ = 0.58). Wolverines avoided human settlements and major roads, preferred moving along forest trails and gentle slopes, and avoided high-altitude deciduous forests. Populations were mainly concentrated in southern Hedong and Qianshao Forest Farms, which are characterized by high habitat integrity, high prey densities, and minimal anthropogenic disturbance. These findings suggest that wolverines may influence boreal trophic networks, especially in areas with intact prey communities, competitors, and spatial refugia from human disturbances. We recommend that habitat protection and management within the natural reserve be prioritized and that sustainable management practices for prey species be implemented to ensure the long-term survival of wolverines. Full article

The increasing threat of quantum computing to traditional cryptographic systems has prompted intense research into post-quantum schemes. Despite SNOVA’s potential for lightweight and secure digital signatures, its performance on embedded devices (e.g., ARMv8 platforms) remains underexplored. This research addresses this gap by presenting the optimal SNOVA implementations on embedded devices. This paper presents a performance-optimized implementation of the SNOVA post-quantum digital signature scheme on ARMv8 processors. SNOVA is a multivariate cryptographic algorithm under consideration in the NIST’s additional signature standardization. Our work targets the performance bottlenecks in the SNOVA scheme. Specifically, we employ matrix arithmetic over ${\mathbb{GF}}_{16}$ and AES-CTR-based pseudorandom number generation by exploiting the NEON SIMD extension and tailoring the computations to the matrix rank. At a low level, we develop rank-specific SIMD kernels for addition and multiplication. Rank 4 matrices (i.e., 16 bytes) are handled using fully vectorized instructions that align with 128-bit-wise registers, while rank 2 matrices (i.e., 4 bytes) are processed in batches of four to ensure full SIMD occupancy. At the high level, core routines such as key generation and signature evaluation are structurally refactored to provide aligned memory layouts for batched execution. This joint optimization across algorithmic layers reduces the overhead and enables seamless hardware acceleration. The resulting implementation supports 12 SNOVA parameter sets and demonstrates substantial efficiency improvements compared to the reference baseline. These results highlight that fine-grained SIMD adaptation is essential for the efficient deployment of multivariate cryptography on modern embedded platforms. Full article

Spatiotemporally mapping winter wheat is imperative for informing and shaping global food security policies. Traditional mapping methods heavily rely on sufficient and reliable samples obtained through labor-intensive fieldwork and manual sample collection. However, these methods are time-consuming, costly, and lack timely and continuous data collection. To address these challenges and fully leverage remote sensing big data and cloud computing platforms like Google Earth Engine (GEE), this paper developed an algorithm for Auto-Generating Winter Wheat Samples for mapping (AGWWS). The AGWWS utilizes historical samples to determine the optimal migration threshold by measuring Spectral Angle Distance (SAD), Euclidean Distance (ED), and Near-Infrared band Difference Index (NIRDI). This facilitates the auto-generation of winter wheat sample sets for the years 2000, 2005, 2010, 2015, and 2021. Approximately two-thirds of the samples were allocated for training, with the remaining one-third used for validating the mapping method, employing the One-Class Support Vector Machine (OCSVM). The Huang–Huai–Hai (HHH) Plain, a major winter wheat production region, was selected to perform the algorithm and subsequent analysis on. Different combinations of the hyper-parameters, gamma and nu, of the OCSVM based on the Gaussian Radial Basis Function Kernel were tested for each year. Following correlation analysis between the winter wheat area derived from the generated maps and the national statistical dataset at the city level, the map with the highest corresponding R² was chosen as the AGWWS map for each year (0.77, 0.77, 0.80, 0.86, and 0.87 for 2000, 2005, 2010, 2015, and 2021, respectively). The AGWWS maps ultimately achieved an average Overall Accuracy of 81.65%. The study then explores the Non-Grain Production of Winter Wheat (NGPOWW) by analyzing winter wheat change maps from 2000–2005, 2005–2010, 2005–2010, and 2015–2021 in the HHH Plain. Despite an overall increase in the total planted area of winter wheat, the NGPOWW phenomena has led to concerning winter wheat planting marginalization. Compensatory winter wheat areas are notably situated in mountainous and suburban cultivated lands with low qualities. Consequently, despite the apparent expansion in planted areas, winter wheat production is anticipated to be adversely affected. The findings highlight the necessity for improved cultivated land protection policies monitoring the land quality of the compensation and setting strict quota limits on occupations. Full article

Siamese network object tracking is a widely employed tracking method due to its simplicity and effectiveness. It first employs a two-stream network to independently extract template and search region features. Subsequently, these features are then combined through feature association to yield object information within the visual scene. However, the conventional approach faces limitations when it leverages the template features as a convolution kernel to convolve the search image features, which restricts the ability to capture complex and nonlinear feature transformations of objects, thereby restricting its discriminative capabilities. To overcome this challenge, we propose replacing traditional convolutional correlation with Slot Attention for feature association. This novel approach enables the effective extraction of nonlinear features within the scene, while augmenting the discriminative capacity. Furthermore, to increase the inference efficiency and reduce the parameter occupation, we suggest deploying a single Slot Attention module for multiple associations. Our tracking algorithm, SiamSlot, was evaluated on diverse benchmarks, including VOT2019, GOT-10k, UAV123, and Nfs. The experiments show a remarkable improvement in performance relative to previous methods under the same network size. Full article

A comprehensive understanding of the consequences of human interactions with mammals is a critical factor in supporting and conserving species in landscapes dominated by humans, which are increasingly threatened. This study aimed to identify the spatial and temporal interactions between humans and mammals. A non-parametric statistical approach with kernel density was used to detect human–mammal temporal interactions. The species interaction factor (SIF) was applied to calculate the spatial overlap based on the two-species occupancy detection model. The activity patterns of medium mammals were nocturnal, diurnal, and cathemeral. The human–medium mammal pairs with SIF values that were <1 and statistically significant included the human–long-tailed macaque (Macaca fascicularis) pair, the human–leopard cat (Prionailurus bengalensis) pair, and the human–barking deer (Muntiacus muntjac) pair. Based on their SIF values and the high overlap in their activity times, the human–macaque pairings had a high risk of conflict. Barking deer and leopard cats displayed a coexistence with humans via time-sharing activities. Due to temporal niche variations with human activities, the existence of nocturnal mammals was relatively uninterrupted. This study showed that most mammals are able to adapt spatially and temporally to various human activities. Nonetheless, efforts to mitigate human–wildlife conflict must be maintained, particularly in the case of severely endangered species, such as the Sunda pangolin. Full article

The development and functional perfection of urban areas have led to increasingly severe fire risks in recent decades. Previous urban fire risk assessment methods relied on subjective judgment, rough data collection, simple linear statistical methods, etc. These drawbacks can lead to low robustness of evaluation and inadequate generalization ability. To resolve these problems, this paper selects the indicator and regression models based on the high-resolution data of the spatial distribution characteristics of Longquanyi distinct in Chengdu, China. and proposes an integrated machine learning algorithm for fire risk assessment. Firstly, the kernel density analysis is used to map the fourteen urban characteristics related to fire risks. The contributions of these indicators (characteristics) to fire risk and its corresponding index are determined by Random Forest (RF), Gradient Boosting Decision Tree (GBDT), and eXtreme Gradient Boosting (XGBoost). Then, the spatial correlation of fire risks is determined through Moran’s I, and the spatial distribution pattern of indicator weights is clarified through the raster coefficient space analysis. Finally, with these selected indicators, we test the regression performance with a backpropagation neural network (BPNN) algorithm and a geographically weighted regression (GWR) model. The results indicate that numerical variables are more suitable than dummy variables for estimating micro-scale fire risks. The main factors with a high contribution are all numerical variables, including roads, gas pipelines, GDP, hazardous chemical enterprises, petrol and charging stations, cultural heritage protection units, assembly occupancies, and high-rise buildings. The machine learning algorithm integrating RF and BPNN shows the best performance (R² = 0.97), followed by the RF-GWR integrated algorithm (R² = 0.87). Compared with previous methods, this algorithm reduces the subjectivity of the traditional assessment models and shows the ability to automatically obtain the key indicators of urban fire risks. Hence, this new approach provides us with a more robust tool for assessing the future fire safety level in urban areas. Full article

Herpes Simplex Virus (HSV) 1 and 2 are persistent infections that affect a significant percentage of United States (US) adults, with 48% having HSV-1 and 12% having HSV-2. Using data stratified by HSV-1 and HSV-2 exposures, this study investigated the association of per- and polyfluoroalkyl substances (PFAS), a group of toxic synthetic organofluorine chemical compounds found in environmental, occupational, and home settings, with allostatic load (AL), an index of chronic physiological stress. Descriptive statistics, multivariable logistic regression, and Bayesian Kernel Machine Regression (BKMR) modeling were used to assess the effects of multi-PFAS exposures on AL using data from the National Health and Nutrition Examination Survey (NHANES) 2007–2014. Results indicated participants not exposed to PFAS exhibited 77% to 97% lower odds of higher AL (p < 0.001). For example, PFOS per unit increase brought forth a 2% odds increase in higher AL (OR: 1.02; 95% CI: 1.00, 1.05; p < 0.05). Participants exposed to PFAS had reduced odds of higher AL (77%–79%), regardless of their HSV-1 and HSV-2 status. PFAS exposure was more prevalent in those with HSV-1 (60%) than in those with HSV-2 (20%) infection, while AL levels were comparable in both groups (17%). BKMR revealed a nonlinear PFAS-AL association and confirmed interactions among PFAS. In summary, PFAS exposure increased the likelihood of higher AL among those with persistent HSV infections. Our study enhances the current understanding of the complex dynamics involving PFAS, persistent infections, and AL, which hold significant implications for public health and clinical intervention strategies. Full article

The growth of rural residential land (RRL) areas has led to the encroachment of cultivated land, which has seriously reduced cultivated land use efficiency (CLUE). This paper takes 42 cities in the middle reaches of the Yangtze River (MRYR) as an example, using the kernel density estimation method, the Super-SBM model, and mediating effect test methods to explore the impact of RRL change on CLUE during 2000–2020. Specifically, based on the analysis of the spatiotemporal distribution characteristics of RRL and CLUE, this paper attempts to further explore the influence path of RRL change on CLUE and test whether there is a mediating effect. The results show that (1) the overall RRL area increased by 30,386.34 hm², except for the decrease in RRL area in a few regions of Hunan Province, and the RRL area in other regions increased. (2) The hot-spot and sub-hot-spot regions of CLUE in the MRYR were mainly concentrated in northwestern Hubei Province and eastern Hunan Province, and the hot-spot and sub-hot-spot regions in Hunan Province are the highest among the three provinces. (3) Under the control of socioeconomic variables, the change in RRL has a significant negative impact on CLUE. (4) The area of cultivated land occupied by rural residential land (CLRRL) has a mediating role during 2000–2020, while the per capita cultivated land area (PCLA) and the rural permanent population (RPP) only have a mediating role during 2000–2010. In the future, the government should strictly prohibit the occupation of cultivated land by RRL and to improve the CLUE. Full article

Machine-learning-based computer vision is increasingly versatile and being leveraged by a wide range of smart devices. Due to the limited performance/energy budget of computing units in smart devices, the careful implementation of computer vision algorithms is critical. In this paper, we analyze the performance bottleneck of two well-known computer vision algorithms for object tracking: object detection and optical flow in the Open-source Computer Vision library (OpenCV). Based on our in-depth analysis of their implementation, we found the current implementation fails to utilize Open Computing Language (OpenCL) accelerators (e.g., GPUs). Based on the analysis, we propose several optimization strategies and apply them to the OpenCL implementation of object tracking algorithms. Our evaluation results demonstrate the performance of the object detection is improved by up to 86% and the performance of the optical flow by up to 10%. We believe our optimization strategies can be applied to other computer vision algorithms implemented in OpenCL. Full article

In order to analyze the spatiotemporal characteristics and space use intensity of residents’ recreational behaviors in the public open space (POS) of communities, the Ruijin, Xincheng, and Fangsong communities, located in the city center, a suburb, and the outskirts of Shanghai, respectively, were studied. The visitor volume data were obtained through the Tencent Travel data. Points of interest of the POS were obtained by the Cat’s Eye app, and the use intensity data were analyzed by kernel density estimation. The use intensity and attraction of POS were verified by stagnation points. The classification of the stagnation points hierarchy showed that the leisure time and distribution of POS visited by residents. There were significantly different in community public spaces, which verifies the rationality of the recreational attraction of the POS. In addition, geodetectors were used to analyze the external factors affecting the characteristics and use intensity of POS. The results show that the percentage of POS area, pedestrian accessibility, population density, percentage of commercial land use, and per-capita occupancy area were important factors, influencing the use intensity of POS, which can provide suggestions for the planning and design of POS in urban communities. Full article

The co-occurrence of felid species in Southeast Asia provides an unusual opportunity to investigate guild structure and the factors controlling it. Using camera-trap data, we quantified the space use, temporal activity, and multi-dimensional niche overlap of the tiger, clouded leopard, Asiatic golden cat, marbled cat, and leopard cat in the Htamanthi Wildlife Sanctuary, Myanmar. We hypothesised that the spatio-temporal behaviour of smaller cats would reflect the avoidance of the larger cats, and similar-sized guild members would partition their niches in space or time to reduce resource competition. Our approach involved modelling single-species occupancy, pairwise spatial overlap using Bayesian inference, activity overlap with kernel density estimation, and multivariate analyses. The felid assembly appeared to be partitioned mainly on a spatial rather than temporal dimension, and no significant evidence of mesopredator release was observed. Nonetheless, the temporal association between the three mesopredators was inversely related to the similarity in their body sizes. The largest niche differences in the use of space and time occurred between the three smallest species. This study offers new insight into carnivore guild assembly and adds substantially to knowledge of five of the least known felids of conservation concern. Full article

The occupancy of residential energy consumers is an important subject to be studied to account for the changes on the load curve shape caused by paradigm shifts to consumer-centric energy markets or by significant energy demand variations due to pandemics, such as COVID-19. For non-intrusive occupancy analysis, multiple types of sensors can be installed to collect data based on which the consumer occupancy can be learned. However, the overall system cost will be increased as a result. Therefore, this research proposes a cheap and lightweight machine learning approach to predict the energy consumer occupancy based solely on their electricity consumption data. The proposed approach employs a support vector machine (SVM), in which different kernels are used and compared, including positive semi-definite and conditionally positive definite kernels. Efficiency of the proposed approach is depicted by different performance indexes calculated on simulation results with a realistic, publicly available dataset. Among SVM models with different kernels, those with Gaussian (rbf) and sigmoid kernels have the highest performance indexes, hence they may be most suitable to be used for residential energy consumer occupancy prediction. Full article

Research on autonomous cars has become one of the main research paths in the automotive industry, with many critical issues that remain to be explored while considering the overall methodology and its practical applicability. In this paper, we present an industrial experience in which we build a complete autonomous driving system, from the sensor units to the car control equipment, and we describe its adoption and testing phase on the field. We report how we organize data fusion and map manipulation to represent the required reality. We focus on the communication and synchronization issues between the data-fusion device and the path-planner, between the CPU and the GPU units, and among different CUDA kernels implementing the core local planner module. In these frameworks, we propose simple representation strategies and approximation techniques which guarantee almost no penalty in terms of accuracy and large savings in terms of memory occupation and memory transfer times. We show how we adopt a recent implementation on parallel many-core devices, such as CUDA-based GPGPU, to reduce the computational burden of rapidly exploring random trees to explore the state space along with a given reference path. We report on our use of the controller and the vehicle simulator. We run experiments on several real scenarios, and we report the paths generated with the different settings, with their relative errors and computation times. We prove that our approach can generate reasonable paths on a multitude of standard maneuvers in real time. Full article

Long-timescale simulations of biological processes such as photosynthesis or attempts to solve NP-hard problems such as traveling salesman, knapsack, Hamiltonian path, and satisfiability using membrane systems without appropriate parallelization can take hours or days. Graphics processing units (GPU) deliver an immensely parallel mechanism to compute general-purpose computations. Previous studies mapped one membrane to one thread block on GPU. This is disadvantageous given that when the quantity of objects for each membrane is small, the quantity of active thread will also be small, thereby decreasing performance. While each membrane is designated to one thread block, the communication between thread blocks is needed for executing the communication between membranes. Communication between thread blocks is a time-consuming process. Previous approaches have also not addressed the issue of GPU occupancy. This study presents a classification algorithm to manage dependent objects and membranes based on the communication rate associated with the defined weighted network and assign them to sub-matrices. Thus, dependent objects and membranes are allocated to the same threads and thread blocks, thereby decreasing communication between threads and thread blocks and allowing GPUs to maintain the highest occupancy possible. The experimental results indicate that for 48 objects per membrane, the algorithm facilitates a 93-fold increase in processing speed compared to a 1.6-fold increase with previous algorithms. Full article

Coexistence between species with similar ecological niches implies species must segregate along one or more niche axes to survive. Space, time, and trophic resources are regarded as the principal axes upon which species segregate. We examined segregation along these niche axes to determine mechanisms underlying coexistence between the two main predators, puma (Puma concolor) and culpeo foxes (Lycalopex culpaeus) in the Andes of Central Chile. We used occupancy modeling to examine space use and overlap, Kernel Density Estimation to determine temporal activity patterns and overlap, and analysis of prey remains in feces to assess diet breadth and similarity. We found high spatial overlap and positive associations between detection of the carnivores lending little support for spatial segregation. Similarly, we found high nocturnal, temporal overlap between pumas and foxes that matched peaks in activity of prey. In contrast, we found relatively low dietary overlap indicating niche segregation likely occurs along the dietary axis. The Puma diet was dominated by introduced, exotic hares and foxes appeared to shift away from hares to rabbits, small mammals, and seeds. Given that lagomorphs are the main dietary resource for pumas in particular, management decisions regarding the control or eradication of such exotic species could negatively affected puma survival. Full article