I am a Machine Learning Engineer at MobileDrive Technology, where I leverage my expertise in computer vision, deep learning, and machine learning. In 2023, I obtained my M.S. degree from the Graduate Institute of Communication Engineering, with a focus on data science, at National Taiwan University, under the guidance of Professor Pei-Yuan Wu. With over three years of research experience, my interests span Deep Learning, Computer Vision, Machine Learning, and Audio Processing.
- Develop an auto-labeling system and build the model for vehicle trajectory planning
- Deploy the trained model to an embedded system, including conversion from Torch to ONNX format and model quantization
- Develop a C++ program for executing model inference on the embedded system
- Improve the existing classification model in the company with about 3% accuracy
- Implement model algorithms for the different tasks, such as recognition, segmentation, and object detection, to Collaborate with colleagues to develop MLops
- Deploy the model as a service using Docker on GCP
- Develop and measure the gaze estimation model using self-supervised learning, Reducing the gaze error from 10 degrees to 1 degree
- Detect the torsional rotation of the eyes using feature matching algorithm
- Write the journal paper with the company
- Full-stack website and system integration
- Deploy the website using Docker and Kubernetes
- Advisor: Prof. Yu-Chiang Frank Wang
- Design and grade homework sets and final project
- Motivate students during T.A. office hours
- Advisor: Prof. Pei-Yuan Wu
- Design the programming exercises
- Advisor: Prof. Jian-Jiun Ding
- Grade homework sets
- Advisor: Prof. Pei-Yuan Wu
- Design and grade the theoretical and programming homework sets
[ICIP 2022] CTGAN: Cloud Transformer Generative Adversarial Network
Github link2022 T-Brain competition - Lung Adenocarcinoma Pathological image segmentation
Private Leaderboard: 2/307, Top1%
Github link2022 AIdea competition - Crops Status Monitoring by Image Recognition
Private Leaderboard: 3/428, Top1%
Github link