A5012132969- Advanced Optimization Algorithms Research
- Matrix Theory and Algorithms
- Robotics and Sensor-Based Localization
- Adaptive Dynamic Programming Control
- Advanced Steganography and Watermarking Techniques
- Thermal properties of materials
- Advanced Neural Network Applications
- Optimization and Variational Analysis
- Video Surveillance and Tracking Methods
- EEG and Brain-Computer Interfaces
- Human Pose and Action Recognition
- Advanced Control Systems Optimization
- Distributed Control Multi-Agent Systems
- Advanced Numerical Methods in Computational Mathematics
- High voltage insulation and dielectric phenomena
- Optimization and Packing Problems
- Computer Graphics and Visualization Techniques
- Control Systems and Identification
- Calcium Carbonate Crystallization and Inhibition
- Advanced Image and Video Retrieval Techniques
- Dielectric materials and actuators
- Advanced Manufacturing and Logistics Optimization
- Chaos-based Image/Signal Encryption
- Advanced Multi-Objective Optimization Algorithms
- Formal Methods in Verification
Guangdong University of Technology
2015-2024
Harbin University of Science and Technology
2023-2024
Hubei University
2024
Dalian University
2023-2024
Dalian University of Technology
2023-2024
University of Science and Technology Beijing
2008-2023
Central South University
2008-2023
GS Caltex (South Korea)
2023
University of Hong Kong
2022-2023
Northeast Petroleum University
2012-2022
Employing attention mechanisms to model both global and local features as a final pedestrian representation has become trend for person re-identification (Re-ID) algorithms. A potential limitation of these methods is that they focus on the most salient features, but may rely diverse clues masked by in different situations, e.g., body, clothes or even shoes. To handle this limitation, we propose novel Salience-guided Cascaded Suppression Network (SCSN) which enables mine integrate into...
Abstract Background Aspartate β-hydroxylase (ASPH) is silent in normal adult tissues only to re-emerge during oncogenesis where its function required for generation and maintenance of malignant phenotypes. Exosomes enable prooncogenic secretome delivering trafficking long-distance cell-to-cell communication. This study aims explore molecular mechanisms underlying how ASPH network regulates designated exosomes program development progression breast cancer. Methods Stable cell lines...
Almost all adversarial attacks in computer vision are aimed at pre-known object categories, which could be offline trained for generating perturbations. But as visual tracking, the tracked target categories normally unknown advance. However, tracking algorithms also have potential risks of being attacked, maliciously used to fool surveillance systems. Meanwhile, it is still a challenging task that on since has free-model target. Therefore, help draw more attention risks, we study algorithms....
This paper aims to achieve high-accuracy 3D object detection, in which we propose a novel Spatial-Channel Attention Network (SCANet), two-stage detector that takes both LIDAR point clouds and RGB images as input generate estimates. The first stage is region proposal network (RPN) put forward new (SCA) module an Extension Spatial Upsample (ESU) module. Using the pyramid pooling structure global average pooling, SCA can not only effectively incorporate multi-scale context information, but also...
Semantic Scene Completion aims at reconstructing a complete 3D scene with precise voxel-wise semantics from single-view depth or RGBD image. It is crucial but challenging problem for indoor understanding. In this work, we present novel framework named Scene-Instance-Scene Network (SISNet), which takes advantages of both in-stance and level semantic information. Our method capable inferring fine-grained shape details as well nearby objects whose categories are easily mixed-up. The key insight...
Recently, a variety of chaos-based image encryption algorithms adopting the traditional permutation-diffusion structure have been suggested. Most these cannot resist powerful chosen-plaintext attack and chosen-ciphertext efficiently for less sensitivity to plain-image. This paper presents symmetric color system based on plaintext-related random access bit-permutation mechanism (PRRABPM). In proposed scheme, new is used shuffle 3D bit matrix transformed from an original image, making RGB...
Frequency selective surfaces (FSSs), also known as spatial filters, are special two-dimensional periodic structures with unique electromagnetic wave transmission and reflection characteristics. With the launch of 6G communication technology research, terahertz band is an important topic technology, miniaturization compactness major features devices. A double-layer ultra-wideband FSS designed, which has a compact cell structure consisting two layers square ring-gap cascades can be used to...
This study was to investigate the effects of housing systems (loose vs. tie stalls) in finishing beef bulls. A total 119 bulls were collected from same farm, with an average live weight 681 kg. The sample comprised 58 loose-housing and 61 tie-stall bulls, each treatment group has six replicates, a fattening period approximately 5 9 months. These animals utilized for comparative research on impact physical activity, blood parameters, animal traits, carcass characteristics, meat quality....
The perception system for autonomous driving generally requires to handle multiple diverse sub-tasks. However, current algorithms typically tackle individual sub-tasks separately, which leads low efficiency when aiming at obtaining full-perception results. Some multi-task learning methods try unify tasks with one model, but do not solve the conflicts in learning. In this paper, we introduce M3Net, a novel multimodal and network that simultaneously tackles detection, segmentation, 3D...
In this paper, we present GaussianPainter, the first method to paint a point cloud into 3D Gaussians given reference image. GaussianPainter introduces an innovative feed-forward approach overcome limitations of time-consuming test-time optimization in Gaussian splatting. Our addresses critical challenge field: non-uniqueness problem inherent large parameter space This space, encompassing rotation, anisotropic scales, and spherical harmonic coefficients, rendering similar images from...
Abstract Epoxy resin, as an adhesive for the main insulation system of large generators, has excellent properties. But its poor thermal conductivity limits capacity enhancement generators. Forming thermally conductive pathways by using electric field to induce high fillers in polymer can effectively improve composite at low loading condition. In this article, a DC is used epoxy resin (EP) two‐dimensional inorganic h ‐BN deflection and head‐to‐tail lap, which make heat flow transfer along...