I am a passionate researcher and developer specializing in smart construction , computer vision, and deep learning technologies. My goal is to bring innovation to the construction industry by leveraging advanced technologies.

Here are some of the projects I have worked on:

vUtils is a custom-built library actively developed by the author, aiming to standardize code modules for various aspects of digital twin technology in construction scenarios. This library plays a crucial role in the author's research projects, where it is extensively used and continuously updated with new features and improvements.

• Multithreaded VideoCapture: Accelerated video capture using multithreading for improved performance.
• VideoCapture with Stabilization: A specialized video capture module designed to remove camera shake in surveillance footage.
• Post-Processing for Semantic Segmentation: Tools for handling and refining the results of semantic segmentation models.

The author is committed to enhancing the functionality of vUtils to better support construction-related digital twin applications, with a strong focus on expanding its capabilities and optimizing performance.

• Description: This repository contains the sample code for my research paper: Semantic segmentation-based framework for concrete pouring progress monitoring by using multiple surveillance cameras(DOI: https://doi.org/10.1016/j.dibe.2023.100283).
• Key Technologies: Semantic Segmentation, Multiple Surveillance, Data Fusion, CRFs.

• Description: This repository is designed to enhance the classification capabilities of the CLIP model through few-shot learning techniques. By leveraging a small number of samples, this project aims to significantly boost the model's performance in classifying worker activities.
• Key Technologies: Few-shot learning, Worker Acitivity Recognition

• Description: This project validates Clip-Adapter on the public CIS dataset in the construction domain, and applied to recognizing rebar tying activities.
• Key Technologies: Few-shot learning, Rebar tying Recognition, CIS Dataset

• Description: This is an example of removing camera shake from fixed surveillance cameras using vUtils.

• CocoDatasetTools: A utility for editing COCO-format annotation files.

• ChessCalibrateTool: Implements Zhang’s camera calibration method.

• CocoSegAdjuster: A tool for refining instance segmentation annotations in COCO datasets.

• FasterVideoCapture: A multithreaded videocapture benchmark.

• Other Projects: ...

Here’s a link to my CV. You can download it to view my professional background, research experience, and academic achievements.

• Construction Automation: Applying advanced technologies like robotics and AI to the construction industry.
• Computer Vision: Enhancing construction site monitoring and safety with real-time visual perception systems.
• Deep Learning: Developing machine learning models to solve complex tasks like semantic segmentation and object detection.

• Email: [email protected]
• GitHub: gaobiaoli

Thank you for visiting my GitHub profile! Feel free to explore my projects, and don't hesitate to get in touch if you have any questions or opportunities for collaboration.