Search

The Chinese Loess Plateau has undergone extensive revegetation to restore degraded land and enhance carbon sequestration. However, soil organic carbon (SOC) sequestrated in the soil profiles of deep-rooted plants has not been fully studied. Here, we investigated the SOC within a 0–23 m profile in farmlands and apple orchards converted from farmlands with different ages (A5, <5 years; A10, ~10 years; A15, ~15 years; A20, >20 years) and the controlling factors on three loess tablelands (Changwu, Qingyang, and Luochuan). The results show that SOC stocks among farmlands and orchards showed no significant difference (p = 0.88); however, SOC stocks showed a trend with tree ages, i.e., a decrease for A5 and A10 but an increase for A15 and A20. For the vertical variability, the SOC stock was the highest within 0–1 m, regardless of the standing age; however, the SOC stock in this layer only accounted for 8.8% of the total SOC stock (97.93 ± 9.18 kg m−2). Climate accounted for 82% of the variations and controlled the changes in SOC in the 0–1 m range, while soil texture dominated the SOC in the soil below 1 m, accounting for 57% of the SOC variations. The variations in SOC in the thick, unsaturated zones provide implications for future land use management and the sustainability of apple orchards in arid regions. [ABSTRACT FROM AUTHOR]

Afforestation is beneficial to improving soil carbon pools. However, due to the lack of deep databases, the variations in soil carbon and the combined effects of multiple factors after afforestation have yet to be adequately explored in >1 m deep soils, especially in areas with deep-rooted plants and thick vadose zones. This study examined the multivariate controls of soil organic carbon (SOC) and inorganic carbon (SIC) in 0–18 m deep under farmland, grassland, willow, and poplar in loess deposits. The novelty of this study is that the factors concurrently affecting deep soil carbon were investigated by multiwavelet coherence and structural equation models. On average, the SOC density (53.1 ± 5.0 kg m−2) was only 12% of SIC density (425.4 ± 13.8 kg m−2), with depth-dependent variations under different land use types. In the soil profiles, the variations in SOC were more obvious in the 0–6 m layer, while SIC variations were mainly observed in the 6–12 m layer. Compared with farmland (SOC: 17.0 kg m−2; SIC: 122.9 kg m−2), the plantation of deciduous poplar (SOC: 28.5 kg m−2; SIC: 144.2 kg m−2) increased the SOC and SIC density within the 0–6 m layer (p < 0.05), but grassland and evergreen willow impacted SOC and SIC density insignificantly. The wavelet coherence analysis showed that, at the large scale (>4 m), SOC and SIC intensities were affected by total nitrogen-magnetic susceptibility and magnetic susceptibility-water content, respectively. The structural equation model further identified that SOC density was directly controlled by total nitrogen (path coefficient = 0.64) and indirectly affected by magnetic susceptibility (path coefficient = 0.36). Further, SOC stimulated the SIC deposition by improving water conservation and electrical conductivity. This study provides new insights into afforestation-induced deep carbon cycles, which have crucial implications for forest management and enhancing ecosystem sustainability in arid regions. [Display omitted] • Soil organic and inorganic carbon varied with land use types and soil depths. • Combined effects of multiple factors on soil carbon were studied at various scales. • Magnetic susceptibility increased the coherence between SOC and total nitrogen. • Water content increased the coherence between SIC and magnetic susceptibility. • We revealed the internal relationship between soil organic and inorganic carbon. [ABSTRACT FROM AUTHOR]

[Display omitted] • The microstructure and mechanical anisotropy of additively manufactured CoCrNi medium-entropy alloys were studied after HIP and UIT. • HIP induces the formation of large number of Cr 2 O 3 particles, and transformation of LAGBs into HAGBs. • UIT contributed to the increase in the number of LAGBs and induced the transformation of columnar grains to equiaxed grains, thus decreasing the anisotropy of AM CoCrNi MEAs. The anisotropy properties of samples in different direction is found in additively manufactured (AM) CoCrNi medium-entropy alloys (MEAs). In this study, the laser direct energy deposition AM CoCrNi MEAs have been subjected to two alternative processing methods: hot isostatic pressing (HIP) and ultrasonic impact treatment (UIT). The effect of HIP and UIT on the microstructure, grain orientation, grain boundary distribution, phase distribution and mechanical properties of AM CoCrNi MEAs were systematically studied. The results show that HIP decreased the partial mechanical anisotropy of AM CoCrNi MEAs, and YS was significantly reduced. The main reason is that the HIP induces the formation of large number of Cr 2 O 3 particles, and transformation of low-angle grain boundaries (LAGBs) into high-angle grain boundaries (HAGBs). Surprisingly, both the yield strength (YS) and ultimate tensile strength (UTS) of the AM CoCrNi MEAs were increased when the UIT was added after every laser deposition of one layer (UIT-1). This is due to the combined effect of quasi-static loading and ultrasonic oscillations of UIT, which results in the dislocation multiplication and the formation of a large number of substructures. In addition, the above strengthening phenomena also lead to decrease partial mechanical anisotropy of AM CoCrNi MEAs. [ABSTRACT FROM AUTHOR]

With the surge in cloud storage popularity, more individuals are choosing to store large amounts of data on remote cloud service providers (CSPs) to save local storage resources. However, users' primary worries revolve around maintaining data integrity and authenticity. Consequently, several cloud auditing methods have emerged to address these concerns. Many of these approaches rely on traditional public-key cryptography systems or are grounded in identity-based cryptography systems or certificateless cryptography systems. However, they are vulnerable to the increased costs linked with certificate management, key escrow, or the significant expenses of establishing a secure channel, respectively. To counter these limitations, Li et al. introduced a certificate-based cloud auditing protocol (LZ22), notable for its minimal tag generation overhead. Nonetheless, this protocol exhibits certain security vulnerabilities. In this paper, we devise a counterfeiting technique that allows the CSP to produce a counterfeit data block with an identical tag to the original one. Our counterfeiting method boasts a 100% success rate ∀ data block and operates with exceptional efficiency. The counterfeiting process for a single block of 10 kB, 50 kB, and 100 kB takes a maximum of 0.08 s, 0.51 s, and 1.04 s, respectively. By substituting the exponential component of homomorphic verifiable tags (HVTs) with non-public random elements, we formulate a secure certificate-based cloud auditing protocol. In comparison to the LZ22 protocol, the average tag generation overhead of our proposed protocol is reduced by 6.80%, 13.78%, and 8.66% for data sizes of 10 kB, 50 kB, and 100 kB, respectively. However, the auditing overhead of our proposed protocol shows an increase. The average overhead rises by 3.05%, 0.17%, and 0.45% over the LZ22 protocol's overhead for data sizes of 10 kB, 50 kB, and 100 kB, correspondingly. [ABSTRACT FROM AUTHOR]

Global warming has accentuated the effects of extreme heat on health. Health insurance, functioning as a risk management tool, has the potential to alleviate these impacts. Consequently, this paper investigates the correlation between extreme heat events and the demand for health insurance in China. Using data from the China Health and Nutrition Survey, we have observed a substantial increase in the likelihood of residents purchasing health insurance during extreme heat events. To be specific, for every extra day of extreme heat events annually, there is a 0.3% increase in the probability of purchasing health insurance. This effect is not uniform across different demographic groups. It is particularly pronounced among middle-aged and elderly individuals, rural residents, those with lower educational levels, higher income brackets, and individuals residing in underprivileged areas with limited access to green spaces and healthcare facilities. Furthermore, our study indicates that the increased frequency of extreme heat events not only impacts individuals' physical health but also triggers negative emotions, which in turn drive risk-averse behavior related to health insurance purchases. These findings carry substantial policy implications for mitigating the economic consequences of climate change. • Extreme heat increases the likelihood of residents purchasing health insurance. • Each additional day of extreme heat boosts health insurance purchases by 0.3%. • This effect varies among different populations. • Mechanistic analyses indicate extreme heat impacting physical and mental health. [ABSTRACT FROM AUTHOR]

Visual behavior intention inference is crucial for enabling escort robots to interact naturally with humans, which is very challenging due to the big inner-class similarity and the small intra-class distinguishability of successive actions in the assistive scenario. To attain a reliable behavior intention inference, not only the current state of behaviors is concerned, but also the semantic information in both spatial and temporal domains plays an important role. This paper presents a segmentation–detection–recognition hierarchical system to represent the spatio-temporal semantic features for formulating descriptions of body parts, trajectories and deep relationships of sub-behaviors. Specifically, a dense trajectory matching scheme based on temporal sampling and Binarized Normed Gradients (BING) algorithm is formulated to segment the 3-Dimensional (3D) behavior cubes, based on which, local trajectories are obtained by clustering dense trajectories according to the distance similarity, and the body parts are then detected by multi-kernel learning of the encoded local features. Moreover, a global three-stream context Convolutional Neural Networks (CNN) is proposed for behavior classification by designing a texture module using expansion, connection and 1D convolution implementations. Based on transfer learning, scene information is also recognized efficiently. Finally, the semantic descriptors are modeled by two cascaded And-Or Graphs (AoGs) constraining the spatial scenarios and temporal sequences. Our unified approach is demonstrated on two public benchmarks containing long-term activities and on an escort robot for real-world applications. • A segmentation–detection–recognition hierarchical framework is proposed to infer behavior intentions for natural human–robot visual interaction. • We design an efficient body part detection scheme by segmenting 3D behavior cubes and fusing local features. • We present a new three-stream context CNN capturing global spatio-temporal relationships of sub-behaviors. • An inference model with two cascaded AoGs constraining spatial scenarios and temporal sequences is constructed. [ABSTRACT FROM AUTHOR]

Laser-arc hybrid welding (LAHW), being a high-efficiency with excellent properties of high welding speed, deep penetration, and good bridging performance, has been paid close attention by domestic and overseas scholars. So far, the lack of suppression and detection of welding defects is still considered the critical technical obstacle that affects its welding quality, particularly for workpieces with industrial requirements. One vital method to conquer this challenging issue is the visual analysis technique with the combination of numerical simulation technique, which has been researched by abundant study outcomes. The primary target of detecting is to collect basic information and to understand the formation mechanism of welding defects. This review firstly describes welding defects online detection technology, such as high-speed image, electrical signal, acoustical signal, and optical signal detection technology. Then much emphasis has been placed on the internal mechanism of forming welding defects, including undercut, humping, porosity, and spatter defects. Particularly, the defect suppression methods are presented in order to restrain and address the welding defect problems. Finally, the current difficulties and potential remedies are discussed to supply an understanding on what still needs to be improved in the LAHW process. This comprehensive review is to offer guidance for those trying to reduce welding defects as they enhance the welding joints' quality. [ABSTRACT FROM AUTHOR]

Even though cetyltrimethylammonium bromide (CTAB) has been successfully used in the surfactant-templating system to create the mesostructured zeolites, the formation mechanism of the pore structure still remains elusive. Here, we investigated the sol-gel and structure evolution in the hierarchical zeolite Beta synthesis focusing on the role of CTAB in a dual templating system. It is found that the induction period was prolonged and the Ostwald ripening was retarded by the presence of CTAB in crystallization. CTAB was involved in the synthesis and incorporated within the solid phase at the first stage when mesopores were formed. However, the incorporated CTAB was released later into the solution when the crystallized microporous structure was developed. The thermogravimetric analysis revealed that CTA+ cation worked both as structure directing agent (SDA) for micropores and mesoporogen for the secondary lager pores in the matrix of hierarchical zeolite Beta. Moreover, CTA+ competed with tetraethylammonium cation (TEA+) to work as the SDA for micropores, with a decrease of the total amount of CTA+ and TEA+ when the concentration of CTAB in the synthesis solution was increased. It is anticipated that the understanding of the role of CTAB in the sol-gel and structure evolution in the hierarchical zeolite Beta synthesis process can provide a basis for designing the synthesis of hierarchical zeolites. [Display omitted] [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel method for dynamic label assignment in temporal action detection (TAD) called Optimal Action Segment Assignment (OASA). The OASA method converts label assignment into an optimal transportation problem by computing the cost matrix between predicted temporal action segments and groundtruths. The unit transportation cost between a predicted temporal segment and a groundtruth pair is defined as the weighted summation of action classification loss and temporal localization loss. Additionally, we deploy Adaptive Estimation of Candidate Segment Number (AE-CSN) to adaptively determine the number of positive samples for each groundtruth. After formulation, the label assignment problem is converted to find a global optimal assignment plan by minimizing the cost. Therefore, OASA eliminates the need for manually designed prior parameters, which exist in fixed label assignment methods, and improves the generalization of the algorithm between different datasets. To evaluate OASA, we also introduce a simple anchor-free temporal action detector called ActionMixer. It consists of two components: Temporal Mixer and Channel Mixer. The Temporal Mixer employs depth-wise convolution layers with large kernels to capture temporal information, while the Channel Mixer mixes and extracts features across the channel dimension. Extensive experiments conducted on the THUMOS-14, ActivityNet-1.3, and EPIC-Kitchens-100 datasets show that ActionMixer equipped with OASA achieves state-of-the-art performance, surpassing other advanced temporal action detection methods. • Optimal Action Segment Assignment is proposed for Temporal Action Detection. • Optimal Action Segment Assignment removes the temporal prior. • ActionMixer is proposed to evaluate Optimal Action Segment Assignment. • ActionMixer achieves state-of-the-art performance on the popular benchmarks. [ABSTRACT FROM AUTHOR]

Local feature matching constitutes the cornerstone of multiple computer vision applications (e.g., 3D reconstruction and long-term visual localization), and has been successfully resolved by detector-free methods. To further improve the matching performance, more recent research has focused on designing sophisticated architectures but endures additional computational overhead. In this study, with a different perspective from previous studies, we aim to develop a deep and compact matching network to improve performance while reducing computing cost. The key insight is that a local feature matcher with deep layers can capture more human-intuitive and simpler-to-match features. To this end, we propose DeepMatcher, a deep transformer-based network that tackles the inherent obstacles of not being able to build a deep local feature matcher with current methods. DeepMatcher consists of: (1) a local feature extractor (LFE), (2) a feature-transition module (FTM), (3) a slimming transformer (SlimFormer), (4) a coarse matches module (CMM), and (5) a fine matches module (FMM). The LFE is utilized to generate dense keypoints with enriched features from the images. We then introduce the FTM to ensure a smooth transition of feature scopes from LFE to the subsequent SlimFormer because of their different receptive fields. Subsequently, we develop SlimFormer dedicated to DeepMatcher, which leverages vector-based attention to model the relevance among all keypoints, enabling the network to construct a deep Transformer architecture with less computational cost. Relative position encoding is applied to each SlimFormer to explicitly disclose the relative distance information, thereby improving the representation of the keypoints. A layer-scale strategy is also employed in each SlimFormer to enable the network to adaptively assimilate message exchange, thus endowing it to simulate human behavior, in which humans can acquire different matching cues each time they scan an image pair. By interleaving the self- and cross-SlimFormers multiple times, DeepMatcher can easily establish pixel-wise dense matches at the coarse level using the CMM. Finally, we consider match refinement as a combination of classification and regression problems and design an FMM to predict confidence and offset concurrently, thus generating robust and accurate matches. Compared with our baseline LoFTR in indoor/outdoor pose estimation, DeepMatcher surpasses it by 3.32%/2.91% in AUC@5 ∘. Besides, DeepMatcher and DeepMatcher-L significantly reduce computational cost and only consume 77.89% and 92.46% GFLOPs of LoFTR. Large DeepMatcher considerably outperforms state-of-the-art methods on several benchmarks, including outdoor pose estimation (MegaDepth dataset), indoor pose estimation (ScanNet dataset), homography estimation (HPatches dataset), and image matching (HPatches dataset), demonstrating the superior matching capability of a deep local feature matcher. • We propose DeepMatcher, a deep Transformer-based network for local feature matching. • We propose SlimFormer to enable DeepMatcher to be extended into dozen layers. • We propose FTM to ensure a smooth transition from CNN to SlimFormer. • We propose FMM to optimize coarse matches, deriving robust and accurate matches. [ABSTRACT FROM AUTHOR]

Local feature matching involves the task of establishing the pixel-wise correspondences between a pair of images. As an integral component of plentiful computer vision applications (e.g., visual localization), this task has been successfully performed using Transformer-based methods. However, these methods typically extract numerous keypoints from sparse texture regions to construct a densely-connected graph neural network (GNN) for long-range feature aggregation, which inevitably triggers redundant message exchange and hampers the learning process. Furthermore, they employ transformer encoder layers that consider images as 1D sequences, leaving them incapable of extracting multiscale local structural information from the images, which is critical for establishing correspondence in image pairs with significantly scales shifts. In this study, we develop FMAP, an innovative detector-free approach that enables accurate local feature matching. For the first issue, FMAP employs an anchor points feature aggregation module (APAM) that captures representative keypoints and discards the extraneous keypoints to build a sparsified GNN for compact yet clean message exchange, with the key insight that the keypoints containing abundant visual information are distinguishable from their neighbors. For the second issue, FMAP proposes a global–local multiscale perception module (GMPM), which incorporates abundant multiscale local context information into global feature representation by employing multiple depth-wise convolutions with varying kernel sizes, thereby generating discriminative features that are robust to scale shifts. In addition, the depth-wise convolutions are utilized in the feed-forward network of the GMPM to further fuse the global context information and local feature representation. Extensive experiments on several standard benchmarks demonstrate that the proposed FMAP method significantly outperforms state-of-the-art methods. Compared to the cutting-edge methods MatchFormer, QuadTree, and TopicFM in relative pose estimation task, FMAP surpasses them by 2.27%, 0.58%, and 1.08% in terms of AUC@5°. Besides, FMAP noticeably outperforms the baseline LoFTR by (2. 38 % , 1. 89 % , 1. 45 %) in terms of AUC@(5°, 10°, 20°). Moreover, we integrate FMAP into an official visual localization framework and conduct a visual localization experiment, with the results showing that FMAP exceeds LoFTR by 2.3% in terms of AP. • We propose a Transformer-based method that achieves excellent matching accuracy. • We constructs sparsified GNN to alleviate redundant message exchange. • We integrate global context and multi-scale local features to boost robustness. [ABSTRACT FROM AUTHOR]