douglas Quan

Key Skills

Software

Appen

Encord

SuperAnnotate

Labelbox

Top Subject Matter

autonomous driving car

NLP in Cantonese

policy text documents classification

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Bounding Box

Classification

Computer Programming Coding

Mapping

Object Detection

Freelancer Overview

My experience as a Research Assistant at the University of Toronto has deeply involved me in the processes of data labeling and AI training, specifically within the realms of language processing and policy analysis. I classified over 300,000 Chinese government policy documents. This project required me to develop a detailed understanding of various industry classification systems, for which I also designed and implemented a method to generate text embeddings, constructing mappings between different patent classification systems. In a prior role with the Department of Linguistics, I was engaged in labeling and data preparation for a Cantonse transliteration project. This involved cleaning, processing and labelling interview transcription data for an RNN model. My work included the development of Python scripts to automate the processing of 63 interview transcriptions. This project not only sharpened my technical skills but also my ability to handle linguistic data sensitively and accurately, ensuring high-quality training data for AI models.

Labeling Experience

Research Assistant

LabelboxImageBounding BoxSegmentation

In my role as a Research Assistant, I was involved in a project dedicated to improving the perception systems of self-driving cars. Our team focused on the accurate labeling of over 100,000 images, which included detecting and annotating various elements such as the direction of cars, other vehicles, pedestrians, and road boundaries. We utilized advanced object detection and semantic segmentation techniques to ensure precise identification of each element. Bounding boxes were applied to vehicles, while polygonal annotations delineated road layouts and pedestrian paths. This meticulous process was crucial in training machine learning models to navigate complex urban environments safely.

In my role as a Research Assistant, I was involved in a project dedicated to improving the perception systems of self-driving cars. Our team focused on the accurate labeling of over 100,000 images, which included detecting and annotating various elements such as the direction of cars, other vehicles, pedestrians, and road boundaries. We utilized advanced object detection and semantic segmentation techniques to ensure precise identification of each element. Bounding boxes were applied to vehicles, while polygonal annotations delineated road layouts and pedestrian paths. This meticulous process was crucial in training machine learning models to navigate complex urban environments safely.

2023 - 2023

Research Assisstant

SuperannotateTextTranslation Localization

Constructed a training dataset for a Cantonese transliteration Transformer Model.

Constructed a training dataset for a Cantonese transliteration Transformer Model.

2021 - 2022

Education

Work History