A5076879802- Design Education and Practice
- Manufacturing Process and Optimization
- Architecture and Computational Design
- Color perception and design
- Product Development and Customization
- Fiber-reinforced polymer composites
- BIM and Construction Integration
- AI-based Problem Solving and Planning
- Semantic Web and Ontologies
- Industrial Technology and Control Systems
- Modular Robots and Swarm Intelligence
- Additive Manufacturing and 3D Printing Technologies
- Collaboration in agile enterprises
- Aesthetic Perception and Analysis
- 3D Surveying and Cultural Heritage
- 3D Shape Modeling and Analysis
- Multi-Agent Systems and Negotiation
- Evolutionary Algorithms and Applications
- Recycling and Waste Management Techniques
- Advanced Numerical Analysis Techniques
- Augmented Reality Applications
- Creativity in Education and Neuroscience
- Building Energy and Comfort Optimization
- E-Learning and Knowledge Management
- Advanced Photocatalysis Techniques
Ningbo University
2025
Sichuan University
2024
West China Hospital of Sichuan University
2024
Southwest Medical University
2024
Nanjing Institute of Technology
2022-2024
Shanghai University of Engineering Science
2021-2024
Affiliated Hospital of Southwest Medical University
2024
Guangxi University
2012-2023
Shenzhen University
2016-2023
Nanchang University
2022
In industry, defect detection involves two kinds of tasks: classification and location, which make it difficult to ensure the accuracy both, also task still challenging in practical application. Based on analysis advantages disadvantages current method, this paper proposes a method based attention mechanism multi-scale maxpooling (MSMP). order effectively improve model, we use Resnet50 as pre-training network construct two-stage model is used be baseline network, introduce MSMP module basis....
In industry, the task of defect classification and localization is an important part detection system. However, existing studies only focus on one it difficult to ensure accuracy both tasks. This paper proposes a system based improved Yolo_v4, which greatly improves ability minor defects. For K_Means algorithm clustering prianchors question with strong subjectivity, Density Based Spatial Clustering Applications Noise (DBSCAN) determine number Anchors. To solve problem low rate small targets...
Changes in clouds can affect the outpower of photovoltaics (PVs). Ground-based cloud images classification is an important prerequisite for PV power prediction. Due to intra-class difference and inter-class similarity images, classical convolutional network obviously insufficient distinguishing ability. In this paper, a method ground-based by improved combined proposed. To solve problem sub-network overfitting caused redundancy pixel information, overlap pooling kernel used enhance...
This study proposes a novel active screw-drive in-pipe robot that can adapt the circular-type and square-type pipe structure. The is composed of four driving units wall-pressing suspension mechanism. Each unit contains motor, transmission train, an electromagnetic brake, which for switching motion route. DC motors drive helical wheels, incline angle wheels be adjusted by using brake. exhibit rolling steering motion. Thus, capable translation movements, rotation screw motions with respect to...
Abstract Excellent recycled carbon fiber (rCF) was obtained using a thermally activated oxide semiconductor process. An alignment device manufactured based on wet‐laid technology. It used to investigate the effect of various nozzle parameters degree rCF mats. The mechanical, electrical conductivity and EMI shielding properties composites reinforced by mats with different layup structures were analyzed. results showed that optimal for aligned included convex shape, diameter 3 mm, channel...
Abstract The two critical issues related to product design exploration are addressed: the balance between stylistic consistency and innovation, control of process under a great diversity requirements. To address these issues, view understanding is first sought. In this view, designs not only categorized as problem-solving activity but also problem-finding activity. A computational framework developed based on it encompasses belief that activities go hand in accomplish tasks an interactive...
Digital holographic microscopy is becoming increasingly useful for the analysis of marine plankton. In this study, we investigate autofocusing and image fusion in digital microscopy. We propose an area metric method improved wavelet-based method. method, a hologram initially segmented into several plankton regions focus plane detection, then applied to these regions. marked map introduced labeling each region with order refocus images that accounts most pixels. The results indicate provides...
The development of a knowledge-based design support system is lengthy and costly process because various computational techniques necessary for intelligent are not readily available in environment. systematisation knowledge needs combined efforts from designers engineers. Existing tools offer limited to which require sophisticated engineering terms representation, inference, control, truth maintenance learning. In this paper, architecture described. existing architectures reviewed first....