A5103096061- Advanced Optical Sensing Technologies
- BIM and Construction Integration
- Photoacoustic and Ultrasonic Imaging
- 3D Surveying and Cultural Heritage
- Additive Manufacturing and 3D Printing Technologies
- Innovations in Concrete and Construction Materials
- Image Processing Techniques and Applications
- Optical Imaging and Spectroscopy Techniques
- AI in cancer detection
- Image Enhancement Techniques
- Hand Gesture Recognition Systems
- Face and Expression Recognition
- COVID-19 diagnosis using AI
- Spectroscopy and Chemometric Analyses
- Image and Object Detection Techniques
- Geotechnical Engineering and Analysis
- Tunneling and Rock Mechanics
- Cell Image Analysis Techniques
- Face recognition and analysis
- Infrared Target Detection Methodologies
- Water Quality and Pollution Assessment
- Wastewater Treatment and Reuse
- Advanced Proteomics Techniques and Applications
- Human Pose and Action Recognition
- Non-Invasive Vital Sign Monitoring
Nankai University
2020-2025
China Agricultural University
2024-2025
PowerChina (China)
2024
Western Sydney University
2023-2024
Powerchina Huadong Engineering Corporation (China)
2024
De novo peptide sequencing directly identifies peptides from mass spectrometry data, playing a critical role in discovering novel proteins and analyzing complex biological samples without reliance on existing databases. To address challenges both speed accuracy, transformer-based model, TSARseqNovo, incorporates two key innovations: Semi-Autoregressive decoder for parallel prediction of multiple amino acids Masking Refinement refining low-confidence predictions. These features significantly...
3D concrete printing (3DCP) attracts significant attention as an innovative manufacturing technology for the construction industry. As one of challenges in 3DCP, failure mechanisms printed structures were not well understood yet and hard to predict. The three-dimensional finite element (FE) method is effective simulate such a layer-by-layer process. However, some existing technical issues FE modelling, including additional initial deformations, identification, selection material models,...
We propose an approach for recognizing the pose and surface material of diverse objects, leveraging diffuse reflection principles data fusion. Through theoretical analysis derivation factors influencing on method concentrates exploits information. To validate feasibility our research, depth active infrared intensity obtained from a single time-of-flight camera are initially combined. Subsequently, these undergo processing using feature extraction lightweight machine-learning techniques. In...
Joint estimation of human body in point cloud is a key step for tracking movements. In this work, we present geometric method to achieve detection the joints from single-frame captured using Time-of-Flight (ToF) camera. Three-dimensional (3D) silhouette, as global feature cloud, extracted based on pre-processed data, angle and aspect ratio silhouette are subsequently utilized perform pose recognition, then 14 derived via features 3D silhouette. To verify method, test an in-house dataset...
A scene plane information recognition method is demonstrated based on data fusion using a single ToF camera. This approach effectively tackles general LiDAR’s deficiencies in identifying planar content, achieving an impressive accuracy of 98.3%.
Proteomics is crucial in clinical research, yet the application of proteomic data remains challenging. Transforming mass spectrometry (MS) into red, green, and blue color (MS-RGB) image formats applying deep learning (DL) techniques has shown great potential to enhance analysis efficiency. However, current DL models often fail extract subtle, features from MS-RGB data. To address this, we developed ProteoNet, a framework that refines analysis. ProteoNet incorporates semantic partitioning,...
In this work, we demonstrate an innovative object detection framework based on depth and active infrared intensity images fusion with a time-of-flight (ToF) camera. A slide window weight (SWWF) method provides fuse image two modalities to localize targets. Then, the information is extracted construct joint feature space. Next, utilize four machine learning methods achieve recognition. To verify method, experiments are performed in-house dataset containing 1066 images, which categorized into...
Ground loss due to earth pressure balance shield tunneling eventually leads a surface settlement which can be an issue of great concern. However, the existing machine learning methods ignore continuous and dynamic nature EPB tunneling. In this work, multivariate time‐series (MTS) model for ground is proposed based on analysis factors processes related combined with characteristics original data involving multiple parameters recorded by machines in real time. A method visualizing MTS features...
This work presents a material identification method based on depth and intensity data fusion from single ToF camera. It addresses the challenge of identifying objects with varying spatial positions surface orientations.
We demonstrate an approach of object identification using a ToF depth camera. Six objects with different materials are recognized data fusion and machine learning methods. A 98.01% accuracy is achieved by KNN algorithm.
With the extracted facial contour from a ToF camera, we demonstrated novel method to recognize whether person is wearing face mask and type using artificial neural network, achieving 97.32% accuracy.