Search

Machine vision based planar grasping detection is challenging due to uncertainty about object shape, pose, size, etc. Previous methods mostly focus on predicting discrete gripper configurations, and may miss some ground-truth grasp postures. In this paper, a pixel-level grasp detection method is proposed, which uses deep neural network to predict pixel-level gripper configurations on RGB images. Firstly, a novel Oriented Arrow Representation model (OAR-model) is introduced to represent the gripper configuration of parallel-jaw and three-fingered gripper, which can partly improve the applicability to different grippers. Then, the Adaptive Grasping Attribute model (AGA-model) is proposed to adaptively represent the grasping attribute of objects, for resolving angle conflicts in training and simplifying pixel-level labeling. Lastly, the Adaptive Feature Fusion and Grasp Aware Network (AFFGA-Net) is proposed to predict pixel-level OAR-models on RGB images. AFFGA-Net improves the robustness in unstructured scenarios by using Hybrid Atrous Spatial Pyramid and Adaptive Decoder connected in sequence. On the public Cornell dataset and actual objects, our structure achieves 99.09\% and 98.0\% grasp detection accuracy respectively. In over 2,400 robotic grasp trials, our structure achieves average success rate of 98.77\% in single-object scenarios and 93.69\% in cluttered scenarios. Moreover, AFFGA-Net completes a grasp detection pipeline within 15 ms.

The purpose of action prediction is to recognize an action before it is completed to reduce recognition latency. Because action prediction has lower latency than action recognition, it can be applied to a variety of surveillance scenarios and responds faster. However, action prediction is more difficult because it cannot obtain the complete action execution. In this paper, we study the action prediction which is based on skeleton data and propose a new network called adaptive graph convolutional network with adversarial learning (AGCN-AL) for it. The AGCN-AL uses adversarial learning to make the features of the partial sequences as similar as possible to the features of the full sequences to learn the potential global information in the partial sequences. Besides, partial sequences with different numbers of frames contain different amounts of information. We introduce temporal-dependent loss functions to prevent the network from paying too much attention to partial sequences whose observation ratios are small, and ignoring partial sequences whose observation ratios are large. Moreover, the AGCN-AL is combined with the local AGCN into a two-stream network to enhance the prediction, proving that the local information and the potential global information in partial sequences are complementary. We evaluate the proposed approach on two datasets and show excellent performance.

Accelerators for machine learning (AI) inferencing applications are homogeneous designs composed of identical cores. Each core, or processing element (PE), contains multiply-and-accumulate units, control logic, and registers for storing and forwarding weights and activations. Testing homogeneous array-based AI accelerator chips by running automatic test pattern generation (ATPG) at the array level results in a high CPU time and pattern count. We propose a constant-testable (C-testable) method for test generation at the PE level such that the ATPG effort does not increase with the number of PEs. Our results show that, compared to the traditional array-level testing, the proposed method achieves up to 4.2× (3.5×), 1530× (2388×), and 170× (142×) reduction in the test pattern count, ATPG runtime, and test cycle count, respectively, for stuck-at (transition) faults in a 256×256 array, while preserving the test coverage. A reconfigurable scan architecture is introduced to enable the proposed C-testable solution for the entire accelerator array. The design-space exploration of a hierarchical test-compaction framework is presented. We also describe four debug solutions for fault localization and diagnosis.

As an important component of intelligent transportation-management systems, accurate traffic-parameter prediction can help traffic-management departments to conduct effective traffic management. Due to the nonlinearity, complexity, and dynamism of highway-traffic data, traffic-flow prediction is still a challenging issue. Currently, most spatial–temporal traffic-flow-prediction models adopt fixed-structure time convolutional and graph convolutional models, which lack the ability to capture the dynamic characteristics of traffic flow. To address this issue, this paper proposes a spatial–temporal prediction model that can capture the dynamic spatial–temporal characteristics of traffic flow, named the spatial–temporal self-attention graph convolutional network (STA-GCN). In terms of feature engineering, we used the time cosine decomposition and one-hot encoding methods to capture the periodicity and heterogeneity of traffic-flow changes. Additionally, in order to build the model, self-attention mechanisms were incorporated into the spatial–temporal convolution to capture the spatial–temporal dynamic characteristics of traffic flow. The experimental results indicate that the performance of the proposed model on two traffic-volume datasets is superior to those of several baseline models. In particular, in long-term prediction, the prediction error can be reduced by over 5%. Further, the interpretability and robustness of the prediction model are addressed by considering the spatial dynamic changes.

Common Portland cement was used as the main matrix material and cementations material, fly ash as admixture, 30 % wt hydrogen peroxide as foaming agent and other additives to prepare cement based condensate porous material. The optimum amount of foaming agent was determined by water absorption and release experiment, and the metal fiber was added to the material to enhance the water collecting performance. As the amount of foaming agent increased, water absorption increased, water release rate decreased, the optimum content of foaming agent was 2.73 %, the material had better water absorption and water release performance, and the addition of metal fiber could obviously improve the water collection performance of cement-based porous condensate material.

Coumarins are generally considered to be produced by natural plants. Fungi have been reported to produce coumarins, but their biosynthetic pathways are still unknown. In this study, Fusarium oxysporum GU-7 and GU-60 were isolated from Glycyrrhiza uralensis, and their antioxidant activities were determined to be significantly different. Abundant dipeptide, phenolic acids, and the plant-derived coumarins fraxetin and scopoletin were identified in GU-7 by untargeted metabolomics, and these compounds may account for its stronger antioxidant activity compared to GU-60. Combined with metabolome and RNA sequencing analysis, we identified 24 potentially key genes involved in coumarin biosynthesis and 6 intermediate metabolites. Interestingly, the best hit of S8H, a key gene involved in hydroxylation at the C-8 position of scopoletin to yield fraxetin, belongs to a plant species. Additionally, nondestructive infection of G. uralensis seeds with GU-7 significantly improved the antioxidant activity of seedlings compared to the control group. This antioxidant activity may depend on the biological characteristics of endophytes themselves, as we observed a positive correlation between the antioxidant activity of endophytic fungi and that of their nondestructively infected seedlings. IMPORTANCE Plant-produced coumarins have been shown to play an important role in assembly of the plant microbiomes and iron acquisition. Coumarins can also be produced by some microorganisms. However, studies on coumarin biosynthesis in microorganisms are still lacking. We report for the first time that fraxetin and scopoletin were simultaneously produced by F. oxysporum GU-7 with strong free radical scavenging abilities. Subsequently, we identified intermediate metabolites and key genes in the biosynthesis of these two coumarins. This is the first report on the coumarin biosynthesis pathway in nonplant species, providing new strategies and perspectives for coumarin production and expanding research on new ways for plants to obtain iron.

Although a few apple (Malus × ×domestica) varieties are self-compatible, little is known about the differences in fruit quality between self- and cross-pollinated apple. In our current study, we compared the fruit quality of self-pollinated apple plants (cultivar 'Hanfu') in self-pollination or cross-pollinated by another cultivar 'Qinguan'. Analysis of fruit quality revealed substantial differences in the external qualities between self- and cross-pollinated apple fruit, but not in the internal qualities. Fruits harvested from self-pollinated 'Hanfu' were smaller and more asymmetrical than those harvested from the cross-pollinated plants. We developed a mathematical model describing how seed number and distribution affect fruit growth. According to this model, the fewer the seeds, the greater the force released from the seeds and the more asymmetrical the fruit. Detection of endogenous hormone and the associated gene expression revealed that gibberellin (GA) levels and GA transporter gene expression on the seedless side were significantly lower than those on the seeded side. Analysis of fruit pectin methylesterase activity and demethylated pectin levels indicated that the lack of GA limits fruit cell wall extension. Additionally, spraying the self-pollinating plants with gibberellic acid increased the fruit weight and lowered the proportion of asymmetrical fruit, recovering the exterior fruit quality to that of the cross-pollinated fruit. Furthermore, exogenous GA treatment increased the wax layer thickness and reduced the fruit water loss rate, leading to a dramatic improvement in fruit storage capacity. Therefore, exogenous GA treatment could be used to ensure regular fruit production of self-pollinated 'Hanfu'.

Glycyrrhiza uralensis Fisch., a well-known medicinal plant, contains flavonoids including liquiritigenin and isoliquiritigenin, and their corresponding glycoside liquiritin and isoliquiritin. Although some genes encoding UDP-glycosyltransferases (UGTs) have been functionally characterized in G. uralensis, other UGTs mechanisms of glycosylation remain to be elucidated. Against this background the aim of the present study included cloning and characterization of two full-length cDNA clones of GuUGT isoforms from the UGT multigene family. These included GuUGT2 (NCBI acc. MK341791) and GuUGT3 (NCBI acc. MK341793) with an ORF of 1473 and 1332 bp, respectively. Multiple alignments and phylogenetic analysis revealed GuUGTs protein of Glycine max had a high homology to that of other plants. Meanwhile, quantitative real-time PCR was performed to detect the transcript levels of GuUGTs in different tissues. The results indicated that GuUGTs was more expressed in roots as compared to the leaves, and significantly up-regulated upon NaCl stress. The recombinant protein was heterologous expressed in Escherichia coli and exhibited a high level of UGT activity, catalyzing formation of isoliquiritin and liquiritin from isoliquiritigenin and liquiritigenin. The key residues of GuUGT2 for liquiritigenin glycosylation (Asn223), isoliquiritigenin (Asp272) were predicted by molecular docking and residue scanning based on simulated mutations. These results could serve as an important reference to understand the function of the UGT family. In addition, the identification of GuUGT2 and GuUGT3 provides a foundation for future studies of flavonoid biosynthesis in G. uralensis.

To achieve the flight mode transfer of the tilt-rotor aircraft, this paper develops a smooth switching method for an weighted L2 robust asynchronously switched system. Considering the asynchronous phenomenon that exists in the switching control of the tilt-rotor aircraft, first, a sufficient condition for the existence of the sub-controller is derived, which guarantees the exponential stability and weighted L2 performance. Besides, the on-line system and the off-line system are separately employed to realize the continuity of the control input signals, which further improve the transient performance and smoothness of the asynchronously switched system. The simulations on the tilt-rotor aircraft verify the effectiveness and applicability of the proposed method.

Global fishery resources have been declining for decades, leading some fisheries to collapse. Although the decline is partly due to man-made chemical contamination, causal chemicals have been identified in only a few cases. We conducted consecutive 3-year investigations of embryonic mortality in Taihu Lake, China, including heavily contaminated northern areas, including Zhushan (ZS), Meiliang (ML), and Gonghu (GH), and the less polluted southeastern Suzhou (SZ). In 2016, 65.8% of crucian carp (Carassius carassius) embryos collected from ZS died before hatching, a substantially higher mortality rate than those observed in ML (21.7%), GH (15.2%), and SZ (2.2%). In 2017, the embryonic mortality rates were 38.8% in ZS, 1.3% in ML, 6.9% in GH, and 3.5% in SZ, and these rates strongly correlated with the concentrations of tris(1,3-dichloro-2-propyl)phosphate (TDCIPP): 104.2, 1.8, 4.6, and 4.1 ng/g lipid weight (lw) in embryos from ZS, ML, GH, and SZ, respectively. In 2018, embryonic mortality decreased to 4.0% in ZS and 1.2% in GH, consistent with decreases in embryonic TDCIPP concentrations to 17.1 and 1.5 ng/g lw, respectively. Moreover, the TDCIPP concentrations in dead embryos (70.5-216.8 ng/g lw) were much higher than those in live embryos (1.2-10.5 ng/g lw). Embryonic mortality was also observed in well-controlled laboratory experiments in which wild crucian carp were exposed to TDCIPP at concentrations similar to those measured in embryos collected from Taihu Lake, thus confirming TDCIPP as a causal factor in mass crucian carp embryo mortality in Taihu Lake. TDCIPP thus poses a threat to the sustainability of fisheries worldwide, given the high worldwide production volume of this chemical and its embryonic lethal toxicity.

In this paper, density functional theory (DFT) is used to study the structure, electron and absorption properties of 6-Amino crystal in the pressure range of 0–300[Formula: see text]GPa. The variations of the lattice constants, bond lengths and bond angles show that they undergo complex transformations under different pressures. More narrowly, the covalent bonds of the benzene ring and the uracil ring in the planar intramolecular structure are broken, and then the new covalent bonds between the adjacent intermolecular structure are reshaped at about 80, 100 and 160[Formula: see text]GPa. The analysis results of band gap and DOS of 6-Amino indicate that the electronic properties of 6-Amino repeatedly transform from the semiconductor system into the metal phase at 80, 180 and 260[Formula: see text]GPa. The absorption spectra show two important structural changes at 100 and 180[Formula: see text]GPa, and present the high optical activity with the change of pressure.

The outbreak of COVID-19 brings huge challenges to the bike-sharing system and even society structure. Thus, it is urgent to fully understand the impacts of pandemic on bike-sharing behavior. This paper proposed a comprehensive approach to investigate the mobility patterns influenced by the outbreak of COVID-19 pandemic with the case in Washington D.C. Multiple-source data, including bike-sharing trip information, COVID-19 information, geographic and POI information, were collected. Although the total bike-sharing trips decreased up to 80% in spatial-temporal analysis, the trips made by casual user still increased. In addition, the docking stations and trips from 2019 to 2021 were utilized to construct the bike-sharing network. The results present that major network properties, such as connectivity, clustering coefficient, and accessibility, experienced significant decrease during the pandemic. Through the detection of community with modularity method, the evolution of community structure before and after pandemic was captured. The increased long-range and long-time bike-sharing trips results in the combination between central communities and outer communities. To better understand the community structure, the POI (Point of Interests) auxiliary analysis was conducted and central community was found to have similar proportion of POIs even during the pandemic. Implications for bike-sharing management and operation policy was also addressed.

Compared with the classic object detection problem, detecting objects in aerial images has some special challenges including huge orientation variations, complicated and large background, and wide multi-scale distribution. Considering these three challenges together, we propose a novel arbitrary-oriented object detection framework consisting of three main parts. Firstly, the Cascading Attention Network (CA-Net) composed of a patching self-attention module and a supervised spatial attention module is proposed for enhancing the feature representations from objects of interest and suppressing the background noises in Feature Pyramid Network (FPN) from coarse to fine. Then, the Adaptive Feature Concatenate Network (AFC-Net) is proposed to adaptively stack the feature maps pooled from all FPN levels as well as the global semantic features, for dealing with the multi-scale change of objects. Lastly, the OBB Multi-Definition and Selection Strategy (OBB-MDS-Strategy) is proposed to regress rotated bounding boxes more smoothly and detect oriented objects more accurately in the training process. Our experiments are conducted on two common and challenging aerial datasets, i.e., DOTA and HRSC2016. Experiments results show that the proposed method has superior performances in multi-orientated objects detection compared with the representative methods.

Zuyi Ma,1,2,&ast; Zhenchong Li,1,3,&ast; Shujie Wang,1,4,&ast; Qi Zhou,5,6,&ast; Zuguang Ma,7 Chunsheng Liu,1,2 Bowen Huang,8 Zehao Zheng,1,2 LinLing Yang,9 Yiping Zou,1,2 Chuanzhao Zhang,1,3,4 Shanzhou Huang,1,3,4 Baohua Hou1,3,4 1Department of General Surgery, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangzhou, 510080, People’s Republic of China; 2Shantou University of Medical College, Shantou, 515000, People’s Republic of China; 3South China University of Technology School of Medicine, Guangzhou, 51000, People’s Republic of China; 4The Second School of Clinical Medicine, Southern Medical University, Guangzhou, 510515, People’s Republic of China; 5Department of General Surgery, Hui Ya Hospital of the First Affiliated Hospital, Sun Yat-Sen University, Huizhou, 516081, People’s Republic of China; 6Department of Liver Surgery, The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, 510000, People’s Republic of China; 7Sanshui Disease Prevention Cure Station, Foshan, 528100, People’s Republic of China; 8Department of General Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100730, People’s Republic of China; 9Guangzhou Medical University, Guangzhou, 511436, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Shanzhou Huang; Baohua Hou Tel +86 13928842869; +86 13609006510Email hshanzh@163.com; hbh1000@126.comBackground: Recent evidence has shown that Solute Carrier Family 39 Member 10 (SLC39A10) promoted tumor progression in several cancer types. The study intended to explore the expression and function of SLC39A10 in hepatocellular carcinoma (HCC).Methods: Multiple bioinformatics analyses were used to evaluate SLC39A10 expression and potential role in HCC. Quantitative real-time polymerase chain reaction and immunohistochemistry were used to confirm SLC39A10 expression. Intro studies were performed to assess the effects of SLC39A10 on HCC cells proliferation and migration. Furthermore, flow cytometry was conducted to identify its specific function in apoptosis of HCC.Results: SLC39A10 was significantly over-expressed in HCC samples from both bioinformatic databases and our cohort. Survival analyses suggested patients with high expression of SLC39A10 had poor overall survival and disease-free survival (P-value < 0.01). Further, the expression of SLC39A10 was positively correlated with tumor-infiltrating lymphocytes and some immune checkpoints like CTLA4, TIM3 and TGFB1. In HCC cell lines, SLC39A10 knockdown inhibited cells proliferation and migration, but promoted apoptosis.Conclusion: An increased SLC39A10 expression was found and served as an unfavorable indicator of survival in HCC. Further studies suggested SLC39A10 promotes tumor aggressiveness and may provide a novel target for HCC therapy.Keywords: Solute Carrier Family 39 Member 10, hepatocellular carcinoma, immune infiltration, prognosis, cancer aggressiveness