Anqi Joyce Yang

I am a Ph.D. student in the Machine Learning Group at the University of Toronto, advised by Prof. Raquel Urtasun. I am also a researcher at Waabi Innovation and a member of the Vector Institute. Previously I completed my M.Sc. in the same group, and my B.Sc. in Computer Science at the University of Waterloo, with a minor in Combinatorics and Optimization.

My general research interests involve computer vision, machine learning and robotics. I have mostly worked with detecting, localizing and reconstructing objects in the self-driving setting using 3D geometry and deep learning.

Email  /  Github  /  Google Scholar  /  Twitter

To tackle LiDAR-based auto-labelling/offboard perception within a two-stage initialize then refine paradigm, we propose LabelFormer, a transformer-based simple, efficient and effective object trajectory-level refinement method.

We introduce a simple yet effective LiDAR-based unsupervised object detection method that uses point clustering and near-range training to bootstrap, temporal consistency to refine, and self-training to iterate.

We present a neural closed-loop sensor simulation method. UniSim turns a single recorded log into a unified LiDAR and camera simulator, which generates realistic sensor data to evaluate autonomy system on safety-critical scenarios.

We leverage class-specific CAD models and in-the-wild sensor data to reconstruct 3D assets with accurate geometry, photo-realistic appearance, part-aware controllability and a representation that is fast to render.

We use learned shape priors to establish "virtual correspondences" and match images with little or no overlaps.

We propose a general framework and a dataset for solving visual Simultaneous Localization and Mapping (SLAM) with multiple asynchronous cameras.

University of Toronto Jan 2021 - Present
Ph.D. Student in Computer Science, Advisor: Prof. Raquel Urtasun

University of Toronto Sept 2019 - Jan 2021
M.Sc. in Computer Science, Advisor: Prof. Raquel Urtasun

University of Waterloo Sept 2015 - Aug 2019
B.Sc. in Computer Science, Combinatorics and Optimization minor

Waabi Innovation Inc. | Toronto, ON, Canada Mar 2021 - Present
Researcher
Full-time researcher in autonomous driving

Uber Advanced Technologies Group | Toronto, ON, Canada Sept 2019 - Feb 2021
Research Scientist
Full-time researcher in autonomous driving, with a focus on localization and mapping systems

Pixar Animation Studios | Emeryville, CA, USA Sept - Dec 2018
Core Engineering Intern
Built post-processing shaders for ambient occlusion, depth of field and motion blur

Uber Advanced Technologies Group | Toronto, ON, Canada May - Aug 2018
Research Intern
Researched on attention-based matching to boost storage efficiency in localization and mapping

Google | Mountain View, CA, USA Jan - Apr 2018
Software Engineering Intern
Key stakeholder to the successful launch of the new Google Code Jam platform

Google | Waterloo, ON, Canada May - Aug 2017

Engineering Practicum Intern
Developed a machine learning system to classify malicious ads with hand-engineered features

Ontario Institute for Cancer Research | Toronto, ON, Canada May - Aug 2016
Software Engineering Intern
Wrote a job-scheduler specific driver for Boutros Lab’s high performance computing interface

Conference and Journal Reviewer: CVPR, ECCV, ICRA, IROS, AAAI, RA-L

© 2021 Anqi Yang. Thanks Jon Barron for the template.