vincentsunnchen /blog vincentschen

vincent sunn chen

vincentsc [at] cs [dot] stanford [dot] edu

I'm a graduate student at Stanford with a concentration in machine learning and a minor in creative writing. I'm interested in shaping datasets to make deep learning more accesible to domain experts in fields like medical imaging.

I'm currently a research assistant in the Stanford AI Lab with Chris Ré, applying Snorkel to computer vision tasks. Most recently, I've trained neural networks on the Autopilot vision team at Tesla. In the past, I worked at Sift Science to fight fraud with machine learning, Xbox to build VR experiences, and EMGuidance to build point-of-care healthcare tools. At Stanford, I had a lot of fun co-directing TreeHacks, the university's flagship, international hackathon.

I also love reading, writing, and photography!

teaching

CS231N: Convolutional Neural Networks for Visual Recognition

Teaching Assistant, Spring 2018

Hosted office hours, advised student projects, and led discussion sections on backpropogation and weak supervision.

projects and papers

Weakly supervised classification of rare aortic valve malformations using unlabeled cardiac MRI sequences

Jason A Fries, Paroma Varma, Vincent S Chen, Ke Xiao, Heliodoro Tejeda, Priyanka Saha, Jared Dunnmon, Henry Chubb, Shiraz Maskatia, Madalina Fiterau, Scott Delp, Euan Ashley, Christopher Ré, James Priest

BioArxiv Preprint 2018.

Automated Training Set Generation for Aortic Valve Classification

Vincent Chen, Paroma Varma, Madalina Fiterau, James Priest and Christopher Ré.

Neural Information Processing Systems 2017, ML4H Workshop. [pdf] [poster]

Using weak-supervision, we learn probabilistic training labels for aortic valve MRIs.

Generating Training Labels for Cardiac Phase-Contrast MRI Images

Vincent Chen, Paroma Varma, Madalina Fiterau, James Priest and Christopher Ré.

Neural Information Processing Systems 2017, Medical Imaging Workshop. [pdf]

Predicting Wealth in NYC from FourSquare Check-ins

Vincent Chen*, Dan Yu*.

CS229: Machine Learning. [blog] [pdf] [code]

We develop a new method to predict demographics based on FourSquare data by engineering features based on check-ins mapped to U.S. census tracts.

Class-conditional Superresolution with GANs

Vincent Chen*, Liezl Puzon*, Christina Wadsworth*.

CS231N: ConvNets for Visual Recognition. [pdf] [poster] [code]

We propose several methods to introduce auxiliary, conditional information into generative adversarial networks (GANs) that produce super-resolution results that better tuned to the human eye.

Sequence-to-Sequence Text Summarization

Vincent Chen*, Liezl Puzon*, Eduardo Torres Montaño*. Advisor: Danqi Chen.

CS224N: Natural Language Processing with Deep Learning. [pdf]

We implement several approaches for sequence-to-sequence summarization on the CNN/DailyMail dataset using attention mechanisms and pointer networks.