A5012178198- Human Pose and Action Recognition
- Hand Gesture Recognition Systems
- Structural Health Monitoring Techniques
- Advanced Measurement and Detection Methods
- Gait Recognition and Analysis
- Seismic Waves and Analysis
- Seismic Performance and Analysis
- Vehicle License Plate Recognition
- 2D Materials and Applications
- Topological Materials and Phenomena
- Photonic and Optical Devices
- Calibration and Measurement Techniques
- Anomaly Detection Techniques and Applications
- Structural Engineering and Vibration Analysis
- Earthquake Detection and Analysis
- Biometric Identification and Security
- Infrastructure Maintenance and Monitoring
- Thermography and Photoacoustic Techniques
- Seismology and Earthquake Studies
- Advanced Neural Network Applications
- Image and Object Detection Techniques
Beijing University of Civil Engineering and Architecture
2024-2025
Civil Aviation Flight University of China
2023-2024
Hebei University of Technology
2023
Huaqiao University
2020-2022
Beijing University of Posts and Telecommunications
2021
This study mainly involves the methods and experiments of using infrared thermal wave imaging detection technology to detect internal defects in aircraft composite materials. The results were discussed analyzed. In this paper, feasibility experiment was verified by simulation. simulation, minimum accuracy detectable is 4 mm radius under mesh division with a correlation coefficient 5. On basis, an accurate method prototype nondestructive testing system for materials based on designed, which...
Topological materials with well-defined surfaces and edges have become a prominent research topic. As topological insulators, MnBi2Te4 thin films, their unique surfaces, exhibit exceptional electron transport properties good applicability in low-noise, high-sensitivity photoelectric detection. This paper reports straightforward, efficient, cost-effective thermal evaporation method for preparing quantum along an investigation into detection performance. These films can be used to fabricate...
Abstract This paper proposes a new neural network learning method to improve the performance for action recognition in video. Most human methods use clip-level training strategy, which divides video into multiple clips and trains feature by minimizing loss function of clip classification. The category is predicted voting from same In order obtain more effective feature, video-level proposed train 3D CNN boost performance. Different with uses as input, entire input. Consistent constraint...
Abstract Nowadays, the security of manhole cover has become a growing concern. In past, manual inspection and Internet Things were often used to ensure safety cover. With advent smart cities, people look forward using cameras as more convenient method detect whether covers are damaged. Based on structure object detection network, we integrate depth information in picture create new model which improves accuracy detecting damaged covers.
In this work, we discuss the feature fusion approach of multisurface skeleton projection images for action recognition. Multisurface are generated from human joint motion trajectories on three surfaces: horizontal-vertical, horizontal- and vertical-depth surfaces. The vision features these contain complementary information different surfaces generally combined by early or late fusion. To learn fuse discriminative each surface effectively, paper proposes a new method called network, which...
In view of the characteristics multiple parameters ZJ119 high-speed cigarette unit, optimal configuration method unit based on materials was studied. Relevant data were collected, collection and processing schemes proposed, correlation between material machine tuning parameters, quality indexes analyzed. The classified by analysis, BP neural network model input for parameter training prediction. After many times research prediction, all that meet requirements index are made into line chart...