The growing explosion in the use of surveillance cameras in public security highlights the importance of vehicle search from a large-scale image or video database. However, compared with person ...re-identification or face recognition, vehicle search problem has long been neglected by researchers in vision community. This paper focuses on an interesting but challenging problem, vehicle re-identification (a.k.a precise vehicle search). We propose a Deep Relative Distance Learning (DRDL) method which exploits a two-branch deep convolutional network to project raw vehicle images into an Euclidean space where distance can be directly used to measure the similarity of arbitrary two vehicles. To further facilitate the future research on this problem, we also present a carefully-organized largescale image database "VehicleID", which includes multiple images of the same vehicle captured by different realworld cameras in a city. We evaluate our DRDL method on our VehicleID dataset and another recently-released vehicle model classification dataset "CompCars" in three sets of experiments: vehicle re-identification, vehicle model verification and vehicle retrieval. Experimental results show that our method can achieve promising results and outperforms several state-of-the-art approaches.
Quantitative reconstructions of the depth to water table (DWT) of ombrotrophic (rain-fed) peatlands are important for understanding the palaeohydrological responses of peatlands to past climate ...changes. This understanding can provide insights into projecting peatlands future variability and evolution. However, the postglacial DWT reconstruction of peatlands in China is challenging due to complications of the atmospheric circulation system, the scarcity of hydrological proxies, and the site-specific nature of hydrological signals. Here we present a postglacial quantitative DWT reconstruction based on the analysis of fossil phytoliths from the Dajiuhu Peatland, central China. The reconstructions were based on a phytolith-DWT calibration model using a weighted averaging partial least-squares regression analysis of peatland topsoil calibration datasets. Three shallow DWT (wet) periods at 13,000–11,500 cal yr BP, 9,600–7,500 cal yr BP, and 3,000 cal yr BP-present, and two extended deep DWT (dry) periods at 11,500–9600 cal yr BP and at 7,500–3,000 cal yr BP are found based on the cluster analysis of phytolith assemblages and reconstructed DWT changes. These five documented hydrological periods are consistent with regional precipitation reconstructions from independent records in the middle Yangtze Valley (MYV). We interpret these changes as mostly reflecting changes in ENSO at various timescales. An amplified ENSO forced a southward Western Pacific Subtropical High and caused the persistence of the Meiyu Front in the mid-lower Yangtze Valley, consistent with the intense rainfall periods in our study region. Our results indicate that phytolith records are a reliable and sensitive proxy for the calibration of water table models and the quantitatively palaeo-DWT reconstruction of peatlands and reveal a remarkable link between the local hydrological variations and the coupled atmospheric-oceanic circulation, which is significant for the prediction of future hydrological changes in the Asian monsoon region under the background of global warming.
•Quantified postglacial DWT of Dajiuhu Peatland based on phytoliths.•The DWT of Dajiuhu Peatland response to the regional precipitation.•Recognized a wetter YD, early and late Holocene and a drier mid-Holocene in the MYV.•Demonstrated a spatial ‘tripole’ mode of palaeohydrology in EASM regions of China.•Proposed the ENSO-related hydrological changes at millennial to annual timescales.
In this paper, we present a large-scale dataset and establish a baseline for prohibited item discovery in Security Inspection X-ray images. Our dataset, named SIXray, consists of 1,059,231 X-ray ...images, in which 6 classes of 8,929 prohibited items are manually annotated. It raises a brand new challenge of overlapping image data, meanwhile shares the same properties with existing datasets, including complex yet meaningless contexts and class imbalance. We propose an approach named class-balanced hierarchical refinement (CHR) to deal with these difficulties. CHR assumes that each input image is sampled from a mixture distribution, and that deep networks require an iterative process to infer image contents accurately. To accelerate, we insert reversed connections to different network backbones, delivering high-level visual cues to assist mid-level features. In addition, a class-balanced loss function is designed to maximally alleviate the noise introduced by easy negative samples. We evaluate CHR on SIXray with different ratios of positive/negative samples. Compared to the baselines, CHR enjoys a better ability of discriminating objects especially using mid-level features, which offers the possibility of using a weakly-supervised approach towards accurate object localization. In particular, the advantage of CHR is more significant in the scenarios with fewer positive training samples, which demonstrates its potential application in real-world security inspection.
In the Jianghan Plain, known as the province of thousands of lakes, monsoon precipitation-induced floods carrying large amounts of sediments can shape the depositional landscape, particularly after ...frequent river flooding. Therefore, the sediments document abundant information on climate change and its interaction with the fluviolacustrine system and paleoflooding. To date, late Holocene paleoflood records are limited, as continuous paleoflood archives are scarce. Here, using grain size end-members and sensitive components in combination with element records from the JH001 sediment core in the Jianghan Plain, nine continuous paleoflood regimes were identified during 3400–3000, 2750–2550, 2050–1820, 1800–1650, 1550–1500, 1350–1100, 1050–1000, 800–750, and 620–600 cal yr BP, respectively. Exploration of the climatic driving factors on flood regimes indicated a good agreement among paleoflood regimes and El Niño-Southern Oscillation (ENSO)-related hydrological records in the Middle Yangtze Valley (MYV) and reconstructed ENSO records. This suggests that ENSO is the dominant influence on flood occurrence in the Jianghan Plain. Strong ENSO activity exerted a southward shift in the subtropical high in the western Pacific and produced a long-lasting rainbelt in the mid-lower Yangtze Valley, which induced the occurrence of floods in the Jianghan Plain. The evolution pattern of sedimentary environment demonstrated that the floods led to river overflow and overbank expansion, even causing the rivers to divert and bifurcate. The overflow channels carried and deposited silt sediments in the floodplain, which shaped the specific alluvial-lacustrine landscape of the Jianghan Plain. Our results indicate that the grain size of core sediments is sensitive to the occurrence of paleofloods, and that ENSO-related paleofloods significantly influenced the fluviolacustrine sedimentary evolution. These findings are highly useful for regional flood planning and environmental protection in the eastern Jianghan Plain in the context of global warming.
•Grain size end-member fractions were sensitive to paleoflood events.•Successive paleoflood records of the Jianghan Plain over the past 4000 cal yr BP were obtained.•The paleoflood events in Jianghan Plain were regulated by ENSO activities.•ENSO-related paleofloods dominated the fluviolacustrine sedimentary evolution.
Person re-identification aims to identify the same pedestrians captured by various cameras from different viewpoints in multiple scenarios. Occlusion is the toughest problem for practical ...applications. In video-based ReID tasks, motion information can be easily obtained from sampled frames, and provide discriminative human part representations. However, most motion-based methodologies are designed for video frames which are not suitable for processing single static image input. In this paper, we propose a Motion-Aware Fusion (MAF) network, aiming to acquire motion information from static images in order to improve the performance of ReID tasks. Specifically, a visual adapter is introduced to enable visual feature extraction, either from image or video data. We design a motion consistency task to guide the motion-aware transformer to learn representative human-part motion information and greatly improve the learning quality of features of occluded pedestrians. Extensive experiments on popular holistic, occluded, and video datasets demonstrate the effectiveness of our proposed method. This method outperforms state-of-the-art approaches by improving the mean average precision (mAP) by 1.5% and rank-1 accuracy by 1.2% on the challenging Occluded-REID dataset. At the same time, it surpasses other methods on the MARS dataset with an improvement of 0.2% in mAP and 0.1% in rank-1 accuracy.
Display omitted
•The hierarchical BiOI/Bi2MoO6 composite prepared by precipitation–deposition method.•The composites showed enhanced visible light activity towards MB and BPA degradation.•The ...BiOI/Bi2MoO6 heterojunction facilitated the separation of electron–hole pairs.•The probable mechanism was proposed to explain the enhanced photocatalytic activity.
A novel hierarchical BiOI/Bi2MoO6 composites are prepared by a facile precipitation–deposition method. The prepared photocatalyst was characterized by X-ray diffraction (XRD), transmission electron microscope (TEM), scanning electron microscopy (SEM), diffuse reflectance spectroscopy (DRS), and X-ray photoelectron spectrometer (XPS). The resulting BiOI/Bi2MoO6 composites exhibit excellent photocatalytic activity and stability towards the degradation of methylene blue (MB) and bisphenol A (BPA) in aqueous solution under visible light irradiation. The optimal composite with 25% BiOI content shows the highest photocatalytic activity for MB degradation. The enhanced photocatalytic activity is mainly attributed to the formation of BiOI/Bi2MoO6 heterojunction that can facilitate the separation and transfer of the photo-generated charge carriers. The roles of active species in the photocatalytic process are discussed by using different types of active species scavengers. Meanwhile, combined with the photoluminescence (PL) and electrochemical impedance spectroscopy (EIS), the degradation mechanism of the photocatalysts is proposed. It is hoped that the work could provide valuable information on the design of specific structure materials with more excellent properties and set the foundation for the further industrial application.
We report a palaeontological record for the last 37.0 cal kyr BP in the Xixi Basin, eastern China, to better understand the correlation between hydroclimatic change and vegetation history during the ...late Pleistocene to early Holocene. Principal component analysis based on the 25 main palynological taxa indicated that dominant taxa fall into three groups: conifers that preferred dry conditions on the left of leading principal component (PC) 1, ferns and wetland communities associated with a humid climate responding to the positive loading of PC1, and Quercus-dominated communities principally corresponding to PC2. Based on this analysis, hydroclimatic conditions were relatively warm/dry during the latter Marine Isotope Stage (MIS) 3 with a high abundance of conifers. This was followed by a generally cold/wet MIS 2 with expansion of ferns and herbs, and a warm/dry MIS 1 with more evergreen species. The sporadic occurrence of evergreen species and drier conditions since about 19.0 cal kyr BP coincides with the beginning of the last deglaciation. An increase in sedimentary accumulation rates and changes in vegetation components at the onset of the Holocene are likely related to maximum transgression. Wet conditions coincide with several cold intervals, including Heinrich events and the Last Glacial Maximum. The overall warm/dry and cold/wet climatic patterns agree well with the palaeoclimatic records elsewhere in the middle Yangtze region, showing similar hydroclimatic forcing mechanisms related to the migration of the Meiyu Front modulated by the El Niño-Southern Oscillation (ENSO) activities. Cold/wet conditions probably correspond to enhanced ENSO and reduced insolation while warm/dry conditions correspond to dampened ENSO and increased insolation. The extremely humid conditions that occur during cold events may result from the combined effects of the Atlantic Meridional Overturning Circulation and ENSO. These results expand our understanding of the monsoonal hydroclimate in eastern China during the late Quaternary period.
•A 37,000-year vegetation is sensitive to the hydroclimatic changes and hydroclimatic history is rebuilt.•Hydroclimate presents a warm/dry and cold/wet pattern.•The long-term hydroclimatic change in eastern China is modulated by ENSO.
Based on linear three-dimensional elasticity theory, the wave equations of coupled Lamb waves in multilayered arbitrary anisotropic composite laminates are derived using a Legendre orthogonal ...polynomial approach. The elastodynamic solution for the propagation of coupled Lamb waves in composite plates is also presented to determine the characteristics of coupled Lamb waves. To verify the applicability and validity of the method, two cases of bi-layered plates formed with isotropic components and anisotropic components, respectively, are primarily manipulated for comparison with earlier known results. Next, the dispersion curves, displacements and stress distributions of Lamb waves in multilayered anisotropic laminates are calculated. The effects of coupling and fiber orientation on the characteristics of the Lamb waves are illustrated. The potential usefulness of the fundamental modes of the coupled Lamb waves is discussed in detail.
•The propagation of coupled Lamb waves is modeled by Legendre polynomial approach.•The applicability and validity of the method are confirmed.•The issue is reduced to solving an eigenvalue problem and the iterative search is avoided.•The coupling and fiber orientation effects are shown and the potential is predicted.
Coronavirus disease 2019 (COVID-19), which is caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has caused a global pandemic. The Omicron variant (B.1.1.529) was first ...discovered in November 2021 in specimens collected from Botswana, South Africa. Omicron has become the dominant variant worldwide, and several sublineages or subvariants have been identified recently. Compared to those of other mutants, the Omicron variant has the most highly expressed amino acid mutations, with almost 60 mutations throughout the genome, most of which are in the spike (S) protein, especially in the receptor-binding domain (RBD). These mutations increase the binding affinity of Omicron variants for the ACE2 receptor, and Omicron variants may also lead to immune escape. Despite causing milder symptoms, epidemiological evidence suggests that Omicron variants have exceptionally higher transmissibility, higher rates of reinfection and greater spread than the prototype strain as well as other preceding variants. Additionally, overwhelming amounts of data suggest that the levels of specific neutralization antibodies against Omicron variants decrease in most vaccinated populations, although CD4
and CD8
T-cell responses are maintained. Therefore, the mechanisms underlying Omicron variant evasion are still unclear. In this review, we surveyed the current epidemic status and potential immune escape mechanisms of Omicron variants. Especially, we focused on the potential roles of viral epitope mutations, antigenic drift, hybrid immunity, and "original antigenic sin" in mediating immune evasion. These insights might supply more valuable concise information for us to understand the spreading of Omicron variants.
PDF
