Currently, I am an Honours BSc student in the Department of Computer Science at the University of Victoria. My interests and enthusiasm are for Computer Vision and Natutal Language Processing for Medicine.

I am exploring how to use medical data and machine learning to help doctors make diagnostic decisions. I am broadly interested in computer vision for improving lesion detection and classification, anatomical structure segmentation and quantification in all type of imaging modalities. I am also fascinated by natural language processing for analyzing genomics and clinical reports, and their linkages between images.

Feel free to shoot me an e-mail if you want to have some serious fun by doing Kaggle or Grand Challenge competitions!

I really hope one day I can push the boundaries of human knowledge.

• bubble_chartUniversity of British Columbia

  assignment_ind Research Co-op

  calendar_today From September 2019 to April 2020.

  location_city Vancouver, British Columbia, Canada

  We are interested in developing machine learning, statistical and signal processing algorithms and software infrastructure to combine various sources of omics and imaging data with major emphasis on discovering novel biological information and biomarkers that can be used for treatment selection in cancer.

  ---

  Artificial Intelligence in Medicine Lab at the University of British Columbia

  Responsibilities include:

  • Researched in patch-level and WSI-level classification for epithelial ovarian carcinoma whole-slide pathology images.
  • Proposed and implemented deep learning-based multi-scale and transfer learning algorithms to improve patch-level classification accuracy, as well as design novel algorithms for slide-level predictions.
  • Benchmarked hand-crafted features, deep learning features and their combinations for patch-level clas- sification.
  • Evaluated the synthetic patch-level pathology images generated by ProGAN as an augmentation step to improve the performance of CNN for patch-level classification.
  • Performed survival analysis using the Cox regression of ovarian cancer patients’ survival data.
  • Designed and built general deep learning for pathology image classification platform using PyTorch, Docker and Kronos.
• bubble_chartUniversity of Victoria

  assignment_ind Research Intern

  calendar_today From May 2019 to August 2019.

  location_city Victoria, British Columbia, Canada

  Computer Vision is used everywhere and is already an essential part of our daily lives. My ultimate goal is to improve the quality of human life through smarter vision systems. Towards this goal, I advance Computer Vision in various levels, mainly through Machine Learning. With the arrival of the Deep Learning revolution, Computer Vision algorithms have become ever more effective, but there is still a lot more in the field to benefit from this advancement. More specifically, my core research interests lie in understanding Geometry, Detecting, Tracking, and Analysing an object of interest. While I approach these problems with Deep Learning and Machine Learning methods, I focus on keeping methods practical, in order for them to be useful in the real-world.

  ---

  Visual Computing Group at the University of Victoria

  Responsibilities include:

  • Researched in patch-level tumour segmentation for the liver hepatocellular carcinoma whole-slide pathology images.
  • Implemented and applied a CVPR 2019 multi-scale with an adaptive weighting deep learning algorithm for automated patch-level detection and segmentation.
  • Developed an unsupervised threshold-based algorithm for segmentation of the tumour area in PET scans as a preprocessing step for image registration.
• bubble_chartConferences

  Conference Name	Role
  Medical Imaging with Deep Learning Conference 2020	Poster Presentation
  Waterloo Mathematics Undergraduate Research Conference 2019	Attendee
  Undergraduate Research Opportunities Conference 2018	Attendee

It is my pleasure to be a co-op student at Kinsol and EncoreFX software development team. I learned so much technical knowledge about software development, testing, and evolution from these eight months co-op and amazing co-workers. But most importantly, the skills of teamwork, interpersonal communication, and problem-solving I learned will benefit me throughout life.

• headsetEncoreFX (now Global Reach)

  assignment_ind Software Developer Co-op

  calendar_today From September 2018 to December 2018.

  location_city Victoria, British Columbia, Canada

  EncoreFX Express is an exciting new offering developed by user interface experts to be both simple and intuitive, and easily accessible by all users no matter their requirements. Express boasts industry leading features like the ability to bulk upload payments and drawdown or pre-deliver on contracts.

  ---

  At EncoreFX, our team of experienced and trusted professionals has a long history of providing foreign exchange risk management solutions and global payment services to businesses worldwide. We are a team of FX Professionals that have joined forces to change the landscape of Corporate FX Risk Management. Our aim is to earn and grow long lasting relationships with our clients by offering industry leading foreign exchange products and services. We are passionate about helping businesses prudently manage their foreign risks and efficiently process global payments.

  ---

  EncoreFX

  Responsibilities include:

  • Developed an online Foreign Exchange Trading and Payment platform, EncoreFX Express, using Angular framework and C# ASP.NET MVC framework.
  • Created user features and interfaces for facilitating interactions, which involves designing, developing, and testing new Angular components on the front-end, as well as building and updating new RESTful API on the back-end.
  • Improved unit testing coverage using Jasmine, developed a Selenium test suite and researched in preventing Cross-Site Request Forgery and Cross-Site Scripting..
• headsetKinsol

  assignment_ind Software Developer Co-op

  calendar_today From May 2018 to August 2018.

  location_city Victoria, British Columbia, Canada

  Kinsol Research Inc. is a provider of research and development services to advanced technology companies. We are a team of researchers and senior engineers specialized in the art of bridging the gap between academia and industry. We excel in innovation, whether it is adapting an existing solution to a particular problem instance or developing a unique approach.

  ---

  Kinsol

  Responsibilities include:

  • Developed several responsive chat bot applications using Python Flask framework, JavaScript ES6, jQuery, and Bootstrap framework.
  • Improved Deep Neural Networks, such as hyper-parameter tuning, regularization and optimization for the chatbot team.
  • Implement the methods which detect and recognize different objects using the YOLO3 algorithm in Python for the traffic analysis team.

"Intellectual growth should commence at birth and cease only at death." - Albert Einstein

I enjoy learning the state-of-the-art algorithms and study from instructors from all over the world. Taking a wide range of courses in different fields is a great way to intake new ideas and skills into my knowledge base.

• all_inclusiveC# Programming for Unity Game Development Specialization

  Unity is my favorite game engine!

  Course Name and Certificate	Date of Completion
  More C# Programming and Unity	October 12, 2018
  Intro to C# Programming and Unity	September 19, 2018

• all_inclusiveDeep Learning Specialization

  First courses in Deep Learning, I take this specialization during the first work term to get prepared for CSC 486B.

  Assignments and Notes can be found in

  Course Name and Certificate	Date of Completion
  Convolutional Neural Networks	July 3, 2018
  Structuring Machine Learning Projects	June 26, 2018
  Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization	June 23, 2018
  Neural Networks and Deep Learning	June 17, 2018

• all_inclusiveData Structures and Algorithms Specialization

  DS&A are arguably the most vital knowledge for computer scienties, it is never enough for me to learn more about DS&A

  Course Name and Certificate	Date of Completion
  Algorithmic Toolbox	August 25, 2017