The increase in the size of the city and the increase in population mobility have greatly increased the number of vehicles on the road, and at the same time brought considerable challenges to the ...traffic management department. In recent years, more and more experts and scholars have devoted themselves to applying sensors, network communication and dynamic adaptive technologies to road traffic management systems. At present, people have not completely overcome all kinds of complex problems in traffic supervision. The complexity of traffic information and the defects of identification algorithms have brought great challenges to intelligent traffic management. This article has launched a research on the intelligent Internet of Things traffic management system, with the detection and recognition of stationary vehicles and seat belts as the key analysis targets. When monitoring stationary vehicles, this paper replaces the background difference algorithm commonly used in dynamic vehicle detection with a new recognition algorithm. From the experimental results, the average detection accuracy of the new algorithm is 96.77% higher than the previous 87.56%. When studying the driver's seat belt detection, this paper combines the YOLOv3 target detection algorithm and the lightweight network structure, and proposes a driver-oriented positioning algorithm. With the increase in the number of lightweight templates, the accuracy of the positioning algorithm has increased from 80.57 to 99.98%. But on the other hand, the detection speed has also changed from 78 to 69 frames/s.
Coastal groundwater is an important resource in the developed region associated with human health and sustainable economic development. To identify the origins of salinity and evaluate the impact of ...water-rock interactions, seawater intrusion (SWI), and evaporation on groundwater in the coastal areas of Zhejiang and Fujian provinces, a comprehensive investigation was performed. Meanwhile, nitrate and fluoride indicators resulting from the anthropogenic activity and SWI were also considered. At last, the water quality index (WQI) of coastal groundwater was evaluated with geochemical and multivariate statistical methods. The results indicated that (1) the groundwater in coastal areas of Zhejiang and Fujian provinces has been affected by SWI to varying degrees. The analysis of selected ion ratios (Na+/Cl- and Br-/Cl-) and isotopic compositions showed that SWI is the predominant cause of increasing salinity in the groundwater of Zhejiang Province, while the cause is water-rock interactions (ion exchange and mineral weathering) in Fujian Province. The hydrochemical evolution path of groundwater in Zhejiang Province is Ca/Mg-HCO3 to Na-Cl, while a different pattern of Ca/Mg-HCO3 to Na (Mg/Ca)-Cl occurs in Fujian Province. However, the trend of SWI development in both provinces was freshening. (2) Nitrification, sewage infiltration, and SWI increased the NO3- content in groundwater. Some of the NO3- concentration in Fujian Province exceeds the standard, and the nitrogen pollution was more serious than in Zhejiang Province. The F- content in coastal groundwater was affected by SWI and mineral dissolution; the F- content in Zhejiang Province was higher than in Fujian Province, which was close to the groundwater standard limit. The average WQI value of Zhejiang was 103.61, and the WQI of Fujian was 61.69, indicating that the coastal groundwater quality in Fujian Province was better than in Zhejiang Province. The results of the study revealed the impact of SWI and anthropogenic activity on groundwater in the southern coastal zone of China and will be valuable for sustainable groundwater resource management.
Groundwater is the primary source of water for domestic use and agricultural irrigation in Jiaodong Peninsula. This study collected 80 groundwater samples from Jiaodong Peninsula to characterize ...groundwater hydrogeochemical processes and the suitability of groundwater for domestic use and agricultural irrigation. The groundwater of Jiaodong Peninsula was categorized as slightly alkaline freshwater, with a Piper diagram classifying most samples as SO4·Cl-Ca·Mg and HCO
3
-Ca·Mg types. Major ions were Ca
2+
, Na
+
, SO
4
2−
, and HCO
3
−
. The major processes driving the hydrochemistry of groundwater were identified as water-rock interactions as well as evaporation. The dissolution of silicate and cation exchange were the predominant hydrogeochemical processes responsible for groundwater chemistry. Four water samples showed seawater intrusion and some indicated pollution from anthropogenic activities such as industry, agriculture, and domestic sewage discharge. Overall, it was found that groundwater in most areas of Jiaodong Peninsula is suitable for domestic use and agricultural irrigation.
Flocculation has a great influence on the biogeochemical cycle by altering the particle size, density and settling velocity of suspended particulate matter (SPM) from coastal to shelf sea areas. ...However, the flocculation process in the shelf sea areas of China has not been systematically studied. In this study, the influencing mechanism of the North Yellow Sea cold water mass (NYSCWM) on the distribution and flocculation process of SPM is studied based on a comprehensive investigation during the summer of 2016. The results revealed that the mass concentration and turbidity of SPM showed a decreasing trend from coastal to offshore areas and an increasing trend from surface to near-benthic layers. Sediments from the Shandong subaqueous clinoform that were resuspended by the tidal mixing effect were the main source of inorganic SPM in the North Yellow Sea in summer. The existence of the bottom cold water mass enhanced the thermocline and pycnocline in the North Yellow Sea shelf area. The pycnocline obstructed the vertical diffusion of nutrients in the NYSCWM, resulting in a subsurface maximum chlorophyll a (chl a) layer, which further contributed to the maximum subsurface volume concentration of SPM. Flocculation was ubiquitous in the study area and was the main reason for the asynchronous variability in the mass concentration, turbidity and volume concentration of SPM. Three different kinds of flocculation mechanisms were identified, including the effects of biological activities in water masses above the NYSCWM, physicochemical effects in the near-benthic nepheloid layer within the NYSCWM, and their combined effects in shallow coastal areas. The flocculation of SPM enhanced the transport of particles from the surface to the bottom of the water column and made an important contribution to the formation of the North Yellow Sea mud deposit.
•The CWM changed the marine physical and biological environments, resulting in variations in SPM flocculation conditions.•Three flocculation mechanisms, dominated by biological activity, physicochemical effects and their coeffects are proposed.•The flocculation processes of SPM play important roles in the formation of mud deposits in the North Yellow Sea.
Neonatal herpes simplex virus type 1 (HSV-1) infections contribute to various neurodevelopmental disabilities and the subsequent long-term neurological sequelae into the adulthood. However, further ...understanding of fetal brain development and the potential neuropathological effects of the HSV-1 infection are hampered by the limitations of existing neurodevelopmental models due to the dramatic differences between humans and other mammalians. Here we generated in vitro neurodevelopmental disorder models including human induced pluripotent stem cell (hiPSC)-based monolayer neuronal differentiation, three-dimensional (3D) neuroepithelial bud, and 3D cerebral organoid to study fetal brain development and the potential neuropathological effects induced by the HSV-1 infections. Our results revealed that the HSV-1-infected neural stem cells (NSCs) exhibited impaired neural differentiation. HSV-1 infection led to dysregulated neurogenesis in the fetal neurodevelopment. The HSV-1-infected brain organoids modelled the pathological features of the neurodevelopmental disorders in the human fetal brain, including the impaired neuronal differentiation, and the dysregulated cortical layer and brain regionalization. Furthermore, the 3D cerebral organoid model showed that HSV-1 infection promoted the abnormal microglial activation, accompanied by the induction of inflammatory factors, such as TNF-α, IL-6, IL-10, and IL-4. Overall, our in vitro neurodevelopmental disorder models reconstituted the neuropathological features associated with HSV-1 infection in human fetal brain development, providing the causal relationships that link HSV biology with the neurodevelopmental disorder pathogen hypothesis.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
In large public places such as railway stations and airports, dense pedestrian detection is important for safety and security. Deep learning methods provide relatively effective solutions but still ...face problems such as feature extraction difficulties, image multi-scale variations, and high leakage detection rates, which bring great challenges to the research in this field. In this paper, we propose an improved dense pedestrian detection algorithm GR-yolo based on Yolov8. GR-yolo introduces the repc3 module to optimize the backbone network, which enhances the ability of feature extraction, adopts the aggregation–distribution mechanism to reconstruct the yolov8 neck structure, fuses multi-level information, achieves a more efficient exchange of information, and enhances the detection ability of the model. Meanwhile, the Giou loss calculation is used to help GR-yolo converge better, improve the detection accuracy of the target position, and reduce missed detection. Experiments show that GR-yolo has improved detection performance over yolov8, with a 3.1% improvement in detection means accuracy on the wider people dataset, 7.2% on the crowd human dataset, and 11.7% on the people detection images dataset. Therefore, the proposed GR-yolo algorithm is suitable for dense, multi-scale, and scene-variable pedestrian detection, and the improvement also provides a new idea to solve dense pedestrian detection in real scenes.
Convolutional neural networks (CNNs) have made significant progress in the field of facial expression recognition (FER). However, due to challenges such as occlusion, lighting variations, and changes ...in head pose, facial expression recognition in real-world environments remains highly challenging. At the same time, methods solely based on CNN heavily rely on local spatial features, lack global information, and struggle to balance the relationship between computational complexity and recognition accuracy. Consequently, the CNN-based models still fall short in their ability to address FER adequately. To address these issues, we propose a lightweight facial expression recognition method based on a hybrid vision transformer. This method captures multi-scale facial features through an improved attention module, achieving richer feature integration, enhancing the network’s perception of key facial expression regions, and improving feature extraction capabilities. Additionally, to further enhance the model’s performance, we have designed the patch dropping (PD) module. This module aims to emulate the attention allocation mechanism of the human visual system for local features, guiding the network to focus on the most discriminative features, reducing the influence of irrelevant features, and intuitively lowering computational costs. Extensive experiments demonstrate that our approach significantly outperforms other methods, achieving an accuracy of 86.51% on RAF-DB and nearly 70% on FER2013, with a model size of only 3.64 MB. These results demonstrate that our method provides a new perspective for the field of facial expression recognition.
Scene text detection is an important research field in computer vision, playing a crucial role in various application scenarios. However, existing scene text detection methods often fail to achieve ...satisfactory results when faced with text instances of different sizes, shapes, and complex backgrounds. To address the challenge of detecting diverse texts in natural scenes, this paper proposes a multi-scale natural scene text detection method based on attention feature extraction and cascaded feature fusion. This method combines global and local attention through an improved attention feature fusion module (DSAF) to capture text features of different scales, enhancing the network’s perception of text regions and improving its feature extraction capabilities. Simultaneously, an improved cascaded feature fusion module (PFFM) is used to fully integrate the extracted feature maps, expanding the receptive field of features and enriching the expressive ability of the feature maps. Finally, to address the cascaded feature maps, a lightweight subspace attention module (SAM) is introduced to partition the concatenated feature maps into several sub-space feature maps, facilitating spatial information interaction among features of different scales. In this paper, comparative experiments are conducted on the ICDAR2015, Total-Text, and MSRA-TD500 datasets, and comparisons are made with some existing scene text detection methods. The results show that the proposed method achieves good performance in terms of accuracy, recall, and F-score, thus verifying its effectiveness and practicality.