Search

Recent two-stage detector-based methods show superiority in Human-Object Interaction (HOI) detection along with the successful application of transformer. However, these methods are limited to extracting the global contextual features through instance-level attention without considering the perspective of human–object interaction pairs, and the fusion enhancement of interaction pair features lacks further exploration. The human–object interaction pairs guiding global context extraction relative to instance guiding global context extraction more fully utilize the semantics between human–object pairs, which helps HOI recognition. To this end, we propose a two-stage G lobal Context and Pairwise-level F usion Features I ntegration N etwork (GFIN) for HOI detection. Specifically, the first stage employs an object detector for instance feature extraction. The second stage aims to capture the semantic-rich visual information through the proposed three modules, Global Contextual Feature Extraction Encoder (GCE), Pairwise Interaction Query Decoder (PID), and Human-Object Pairwise-level Attention Fusion Module (HOF). The GCE module intends to extract the global context memory by the proposed crossover-residual mechanism and then integrate it with the local instance memory from the DETR object detector. HOF utilizes the proposed pairwise-level attention mechanism to fuse and enhance the first stage's multi-layer feature. PID outputs multi-label interaction recognition results with the input of the query sequence from HOF and the memory from GCE. Finally, comprehensive experiments conducted on HICO-DET and V-COCO datasets demonstrate that the proposed GFIN significantly outperforms the state-of-the-art methods. Code is available at https://github.com/ddwhzh/GFIN. [ABSTRACT FROM AUTHOR]

To enhance vibroacoustic radiation and propagation predictions, a novel integrated computation model (ICM) combining the finite element method (FEM), equivalent source method (ESM), and normal mode method (NMM) is proposed. This model overcomes the complex computation domain (CCD) issue inherent in current ICMs, improving accuracy and reducing model complexity. Initially, the FEM is used to compute the vibracoustic radiation from the volume source. Subsequently, the ESM is introduced to estimate Green's function within the near field. Additionally, the fast multipole method (FMM) is introduced to streamline multipoint interaction to that between two elements and enhance the computation speed of the ESM. Furthermore, the NMM is used to compute Green's function in the far field. In addition, the Hamiltonian analysis method (HAM) is employed to search for broadband wave numbers of the NMM and facilitate fast solutions for far field propagation. The simulation results indicate that the FEM-ESM-NMM model effectively addresses the CCD issue encountered with current ICMs. Moreover, computation, once accelerated, has been reduced by 40% in comparison to the ESM and NMM, and has seen an 80% reduction relative to the FEM. • Examines the issue of complex computation domain in integrated models from a distance-based perspective. • Presents an integrated model efficiently calculating underwater target sound fields at all distances. • Optimization algorithms were introduced to significantly improve the computational efficiency of the integrated calculation model. [ABSTRACT FROM AUTHOR]

The development of a photocatalyst with high activity, large specific surface area and precise band gap adjustment is the focus of the field of photocatalysis in realizing energy conversion and improving environmental issues. In this paper, a simple synthesis process for co-doping of KCl and NH 4 Cl to precisely adjust the band gap of g-C 3 N 4 was developed. Various characterization results show that doping with KCl and NH 4 Cl can reduce the band gap of the carbon nitride, increase its visible light absorption, improve its carrier separation efficiency, thus showing good photocatalytic activity. Cl and K enter the g-C 3 N 4 layers via intercalation and form a dual-element charge transport channel between these layers, thereby facilitating the transport of electrons. The introduction of NH 4 Cl increases the specific surface area of the composite, thus providing abundant adsorption and reaction sites. The co-loaded photocatalyst CN-NH 4 Cl-0.05KCl was used to degrade rhodamine B (RhB) and tetracycline (TC) and showed excellent photocatalytic activity. A possible mechanism of photocatalytic degradation is proposed based on the analytic results. ▪ • The band gap of g-C 3 N 4 is adjusted by doping with KCl and NH 4 Cl to suppress carrier recombination. • The introduction of NH 4 Cl into g-C 3 N 4 is conducive to the formation of mesoporous channels. • Cl and K enter the g-C 3 N 4 layers via intercalation and form a charge transport channel between these layers. • The co-loaded photocatalyst CN-NH 4 Cl-0.05KCl has an excellent photocatalytic performance. • Possible photocatalytic mechanisms are proposed on the basis of the analytic results. [ABSTRACT FROM AUTHOR]

Coupling agents are widely used in order to improve the properties of electrically conductive adhesives (ECAs). The effects of different coupling agents on the characteristics of epoxy-based ECAs were investigated in terms of thermal, rheological, electrical and mechanical properties. Results showed that the ECAs had favourable flow, adhesion and good conductivity. In particular, the ECAs filled with Ag flakes treated with the silane coupling agent 3-aminpropyltriethoxysilane (KH-550) exhibited the lowest electrical resistivity (5.22 × 10−4 Ω cm) and the highest shear strength (20.32 MPa) among the ECAs. Besides, it had the minimum viscosity 154 Pa s at 5 rpm/25 °C. It was also found that the adding of coupling agents had different effects on the aging electrical and shear strength of the ECAs. According to this study, the use of KH-550 had a significant electrical conductivity improvement and shear strength increase before and after aging. In addition, a strong adhesion/bonding between the conductive fillers and the resin matrix was revealed by scanning electron microscopy studies. [ABSTRACT FROM AUTHOR]

In the scenario of UAC(Underwater Acoustic Communication) between the USV(Unmanned Surface Vehicle) and underwater node, the affection originated from sea surface random fluctuation to UAC system stability cannot be neglected due to the small size of hull and the shallow draft, especially when USV is sailing near sea surface. The experimental results indicate that the random fluctuation of the sea surface has several influences both on signal propagation and multi-channel structure, and they are difficult to describe quantitatively. Following the research of?marine environment – channel characteristics – underwater acoustic communication', this paper proposes a shallow sea wave spectrum model(USV_SW) to design UAC system based on measured data including the environment parameters, wind speed, wind direction, and real-time water depth. Moreover, the verification of elevation is presented between the proposed USV_SW model and CORS_GPS(Continuously Operating Reference Stations GPS) data collected on USV. Therefore, we calculated the average acoustic reflection coefficient of random fluctuation interface which is applied for the following simulation analysis of vertical and horizontal channels influenced by interface fluctuation. Simulation results show that the horizontal channel is affected more by random interface fluctuation and its characteristic signal space–time fluctuation is noteworthy in bad sea conditions. The analysis of the random fluctuation interface on the channel structure provides data and theoretical support for the design of the USV underwater acoustic communication scheme. [ABSTRACT FROM AUTHOR]