Education
PhD in Computer Science
University of Toronto, 2029
B.A.Sc in Mechatronics Engineering
University of Waterloo, 2023
Experience
University of Toronto - Research Associate
July 2021 - August 2024
- Researched in robot vision and manipulation toward self-driving labs supervised by Prof. Alán Aspuru- Guzik, Prof. Florian Shkurti, and Prof. Animesh Garg.
- Constructed a general-purpose robotic framework by integrating language planner, perception and robot control that is capable with conducting electrochemical experiments without human intervention.
- Developed a novel multiview-based transparent object pose estimation method that achieves the state-of-art performance and enables robot to manipulate transparent vessels for chemistry experiments.
Huawei Technologies Co., Ltd. - Computer Vision Research Intern
September 2020 - December 2020
- Led research in self-supervised representation learning and class-agnostic counting, and co-authored the paper about contrastive learning. Mentored by Dr. Hailin Hu and Yuhao Zhou.
- Designed and built the new framework based on supervised contrastive methods that pretrained backbones to improve performance on downstream tasks and alleviate relies on labeled data.
Ford Motor Company - Firmware Developer
May 2021 - August 2021
- Developed code in C for bootloader and BSP on Qualcomm 5G chips used in the telematic control unit (TCU) of vehicles.
- Designed the token verification pipeline for new version of TCU that secures devices from being hacked and flashing images without authorization.
- Supported image flashing, partition and bootchain updates on NAND-based devices so that OTA update can perform smoothly.
Miovision Technologies Inc. - Computer Vision Engineer
January 2020 - May 2020
- Built the first fisheye semantic segmentation dataset from scratch with custom annotation pipeline using AWS and Python libraries.
- Researched in modern semantic segmentation models, conducted experiments using fisheye images, and delivered reports on different architectures for further research.
- Implemented and trained the DeepLabV3+ model in TensorFlow as proof-of-concept to perform scene understanding on traffic intersection.
Eagle Vision Systems - Deep Learning Engineer
July 2019 - January 2020
- Designed and built a convolutional neural net using Keras and OpenCV in Python to classify garbage bins by colors and achieved 98.8% accuracy.
- Wrote scripts that convert input data to TFRecord format and implemented TensorRT to increase the processing speed by 30%.
HiRide Share Ltd - Full Stack App Developer
July 2019 - January 2020
- Led a group of two co-op students to build the front-end for an in-app messaging feature that allows real-time messaging between users using React Native, Redux, and Socket.io.
- Built a new screen for a navigation feature that renders users’ information, fetches data from third party APIs, shows their real-time location and navigates drivers using Socket.io.
- Designed and developed a new RESTful API for messaging and navigation features that can insert, update, and remove data in MongoDB using Node.js, Express.js, and Mongoose.
