Events Calendar

ECE/ME Seminar: Wei-Lun Chao

Title: Perception and Learning with Imperfect Data: Insights from the Real World

Abstract: Recent advances in computer vision and perception have largely relied on large-scale, high-quality data. However, in many real-world scenarios, the vast diversity of objects, environments, and tasks makes it challenging to collect sufficient data to build generalizable models. In robotic applications such as self-driving cars, ambiguous perceptions due to occlusions or sensor limitations further complicate the problem. These challenges cannot be solved by simply scaling up data collection. Instead, we need algorithms capable of perceiving and learning from imperfect data.

In this talk, I will share key milestones from my research on perception and learning with imperfect data. I will highlight solutions that incorporate real-world insights to introduce effective inductive biases and auxiliary information, mitigating data imperfections. Specifically, I will present a series of works in self-driving perception that address challenges related to sensor limitations, labeled training data, and diverse driving scenarios. Additionally, I will introduce my recent research in a distinct domain—visual recognition for biological organisms—where the extreme fine-grained nature of trait segmentation and the development of vision foundation models for the Tree of Life demand efficient solutions with limited data. I will conclude with a discussion on broader implications and future directions in learning with imperfect data across multiple domains.

Bio: Wei-Lun (Harry) Chao (https://cse.osu.edu/people/chao.209) is a Distinguished Assistant Professor of Engineering Inclusive Excellence in the Department of Computer Science and Engineering (CSE) at The Ohio State University (OSU). His research focuses on machine learning and computer vision, with applications spanning visual recognition, autonomous driving, biology, and healthcare. He aims to develop fundamental understandings and robust, widely applicable algorithms to tackle real-world challenges. He is particularly interested in learning with imperfect data, including limited, noisy, heterogeneous, distribution-shifting, and inaccessible data. His contributions have been recognized with several awards, including the OSU College of Engineering Lumley Research Award (2023), the OSU CSE Faculty Teaching Award (2024), and the CVPR 2024 Best Student Paper Award. Before joining OSU in 2019, he was a Postdoctoral Associate at Cornell University (2018–2019), working with Kilian Weinberger and Mark Campbell. He earned his Ph.D. in Computer Science from the University of Southern California (2013–2018) under the supervision of Fei Sha.