About me

Hello! I am currently in my last year of undergraduate studies at the University of Toronto studying Computer Science with a specialization in Computer Vision. My interests lie at the intersection of robotics and machine learning. When I am not working you can find me training Brazilian Jiu-Jitsu, running, or photographing abandoned buildings.

Publications

* indicates equal contribution.
Large Language Models for Chemistry Robotics
Naruki Yoshikawa, Marta Skreta, Kourosh Darvish, Sebastian Arellano-Rubach, Zhi Ji, Lasse Bjørn Kristensen, Andrew Zou Li, Yuchi Zhao, Haoping Xu, Artur Kuramshin, Alán Aspuru-Guzik, Animesh Garg, Florian Shkurti,
Autonomous Robots (2023)
[paper] [project page]
Chemistry Lab Automation via Constrained Task and Motion Planning
Naruki Yoshikawa, Andrew Zou Li, Kourosh Darvish, Yuchi Zhao, Haoping Xu, Artur Kuramshin, Alán Aspuru-Guzik, Animesh Garg, Florian Shkurti,
Under Review
[paper] [project page]

Experience

Sanctuary AI

March 2024 - Present

I am currently working at Sanctuary AI as a Junior DataOps Engineer.

Sanctuary AI

Summer 2022 Summer 2023

I worked at Sanctuary AI as a Software Co-op on the Machine Learning team. Summer 2022 I worked on object grasping and manipulation using a multi-fingered robotic hand. I implemented a pipeline that would generate and evaluate possible grasps for known objects. Summer 2023 I designed and developed a real-world SLAM data collection and annotation pipeline.

Robot Vision and Learning Lab

Summer 2021 Spring 2023

I was a research assistant intern at the RVL Lab supervised by Florian Shkurti. We worked on stochastic planning for autonomous surface vessel (ASV) navigation using satelite images. I helped design a software stack that supported communication between ROS, the web front-end visualization, and the Gazebo simulation or real-world.

University of Toronto

January 2021 - May 2023

I was a Teaching Assistant at the Department of Computer Science at the University of Toronto. I have helped teach the courses CSC207H1: Software Design and CSC209H1: Software Tools and Systems Programming. My responsibilities include hosting weekly labs for students, grading, providing the instructor with feedback on course material and student evaluations.

Spring Air Systems

May 2021 - September 2021

I was a Software Developer Intern at Spring Air Systems. I developed new features and resolved existing problems in the web application (.NET framework) with SQL and an AutoCAD schematic automation system. Created new and modified existing unit tests for the web application and AutoCAD system. Refactored existing code to be more modular and extensible for future use.

Projects

Vehicle Detection and Motion Forecasting

[Final Project], University of Toronto, CSC498: Making Your Self-driving Car Perceive the World. With C. Zhang.

[report 1] [report 2] [code]

Models for vehicle detection and motion forecasting on the PandaSet dataset. Implemented preprocessing by creating BEV voxelized images from LiDAR data. Developed a heat-map object detection model using focal loss and anisotropic gaussians. Built a model that predicts a parametric distribution over future trajectories.

Hand Drawing Generative Model

[Personal Project]

[report] [code]

PyTorch implementation of the cDCGAN for Google’s ‘Quick, Draw!’ dataset. Tested various model architectures for better results. Implemented techniques such as label noise and instance noise for a more stable model. Visualized results and training data over time to better tune model hyperparameters.

Evaluation of Sound Enhancing GANs

[Final Project], University of Toronto, CSC413: Neural Networks and Deep Learning. With S. Das and P. Bajaria

[paper]

Wrote a mock NeurIPS paper as a final group project. The project focused on using the GAN architecture for audio synthesis. For this paper, my group and I examined the performance of the vanilla Speech Enhancement GAN against variations that include the Departure From Normality (DFN) metric and Wasserstein loss (WSEGAN).

Hobbies

Brazilian jiu-jitsu

I have been training Brazilian jiu-jitsu for about 8 months in both Gi and No-Gi disciplines. Currently I train mostly at the University of Toronto Grappling Club.

Running

Running has been a big part of my life for about the past 7 years. My best half-marathon time is 1:43:43. I am training for the upcoming Toronto Waterfront Marathon race in October.

Photography

I have recently picked up photography. Specifically I combine photography with urban exploration. You can find my work on instagram. I use the Canon EOS Rebel T7 camera.