Kevin Tan

I am a second-year Masters student at Stanford University in the Department of Symbolic Systems. I currently work in the Stanford Vision and Learning Lab (SVL) under the supervision of Prof. Juan Carlos Niebles and Prof. Fei-Fei Li.

Before coming to Stanford, I did my undergrad at UC San Diego where I was fortunate to work with Prof. Zhuowen Tu and supported by the Halicioglu Data Science Institute Scholarship. Over the summers, I've also spent time at Arctype Inc as a ML Engineer and IBM Accessibility Research as a Research Intern.

kevintan [at] cs [dot] stanford [dot] edu

Email  /  CV  /  Twitter  /  LinkedIn  /  Github

• [Sept 2021] My thesis is available here.
• [June 2021] I have graduated from Stanford. I am seeking a fulltime research engineering position.

My research interests lie at the intersection of computer vision, machine learning, inverse graphics, and cognitive science. I draw inspiration from human cognition to build machines capable of learning, perceiving, and understanding the complex visual world just as humans can do so effortlessly.

We propose a model with Intra- and Inter Transformers, enabling joint spatial and temporal reasoning on multiple visual concepts of objects, relationships, and human poses.

We introduce a novel framework that leverages a joint layout-texture embedding space for controllable image synthesis from scene graphs.

Stanford

We propose an encoder-decoder method for scene graph generation with image reconstruction as a self-supervisory signal.

UCSD

Inspired by how humans anticipate future scenes, we explore a self-supervised spatiotemporal method for generating future frames in videos.

We exploit structure in Chinese characters by learning embedding weight matrices for word-level, character-level, and radical-level for improved neural machine translation.

Implementation of coordinate descent with Gauss-Southwell selection rule and backtracking line search on the UCI Wine dataset.

UCSD

• COGS118B: Introduction to Machine Learning (Fall 2018)

• My favorite books are The Three-Body Problem, Flowers for Algernon, and Ready Player One.