Mobile +65 86986465
Email: [email protected]
LinkedIn: https://www.linkedin.com/in/zhiyuan-li-a19b0863/
Nationality: Singaporean
Dedicated and highly motivated machine learning engineer with proficiency in programming languages such as Python and Java and experience in using deep learning frameworks such as PyTorch and TensorFlow. Experience in optimizing model through TensorRT. Knowledge in video data pre-processing techniques such as intrinsic camera calibration, and image warping through homographic matrix. Proven track record of producing accurate and effective models using different data types, including text, image, video, and physiological signals. Bilingual, organized, and adaptive. A team player who contributes to innovation, and advancements of machine learning as well as creating solutions that brings convenience to people and society.
Project summary:
- Covid-19 Fake News Detection NLP project
- Stress Detection using Conventional Machine Learning
- Violence and Weaponized Violence Detection using Energy Efficient Deep Neural Networks
- Emotion Detection in Video Conferencing Applications
- End-to-end Deep Neural Network for Multimodal Biometric Authentication and Driver Stress Detection.
Python
PyTorch
TensorFlow
TensorRT
Image/Video Classification
Text Classification
Scikit-learn
Probability and Statistics
Data Preprocessing
Object Detection
Signal Classification
Image Segmentation
Convolutional Neural Network
Linux Environment
Data Structure and Algorithms\
Machine Learning, Senior Engineer
ST Engineering
Singapore
Jul 2023 – Present
- Worked on projects such as Video Anomaly Detection, Unattended Object Detection, Workplace Safety, and Visible and Thermal Image Fusion
- Optimized the models through TensorRT and torch2trt
- Performed data pre-processing such as Intrinsic camera calibration, and image warping through homographic matrix
- Improved image processing speed by 20 times through CUDA parallel programming using PyCUDA for customized cv2 functions
Software Engineer
NCS Pte Ltd
Singapore
Jul 2019 – Nov 2020
- Setup TYK API Gateway through Amazon Elastic Container Service (ECS) and Elastic Compute Cloud (EC2) for People’s Association and MuleSoft API for Singapore Tourism Board
- Transformed older version API to TYK API and onboarded the API for the Ministry of National Development
- Developed and maintained Digital Notice System (tax letters) using Java for the Inland Revenue Authority of Singapore
Deep Learning Intern
ASM Pacific Technology
Singapore
Feb 2018 – Jul 2018
- Implemented and trained deep learning model using TensorFlow and Masked R-CNN for segmentation and classification of defective semi-conductor epoxy resins
- Labelled and collected image data of the semi-conductor epoxy resins
Chatbot Intern
MoneySmart Group
Singapore
May 2017 – Jul 2017
- Developed and maintained chatbot using python and Flask
- Designed and managed database using MySQL and PeeWee
Mohamed bin Zayed University of Artificial Intelligence
Master of Science
Adu Dhabi
Jan 2021 – Dec 2022
- Machine learning Specialization
- Full scholarship
- Thesis topic: End-to-end Deep Neural Network for Multimodal Biometric Authentication and Driver Stress Detection (Project aims to replace physical keys with owner’s biometrics to provide added security for vehicle access and to monitor driver’s stress level so as to prevent road rage accidents. Application in real-world for fraud detection)
- GPA 3.84 / 4
Mohamed bin Zayed University of Artificial Intelligence is the first graduate-level, research-based academic institution in the world, specializing in machine learning, computer vision and natural language processing
National University of Singapore
Bachelor of Engineering (Honours) Computer Engineering
Singapore
Aug 2015 – Jul 2019
- Thesis topic: Clothes Classification and Recommendation (Project aims to identify the type of clothes present in the image and recommend consumers related clothing types from online stores. Application for customer experience using Virtual Reality and Augmented Reality)
- GPA 3.5 / 5
Professional Qualifications / Certifications
- Professional Scrum Master 1
Second position in Innovation Challenge for Entrepreneurship (ICE) 2022, Abu Dhabi
- Competition across all GCC countries
- Developed a prototype for classifying student’s emotions through real-time analysis of video conferencing applications such as Zoom and Webex and provide feedback to teaching instructors for immediate intervention.
Mode | Language | Proficiency |
---|---|---|
Spoken | English Mandarin |
Native Native |
Written | English Mandarin |
Competent Competent |