Yuan-Hong Liao

University of Toronto, Vector Institute

prof_pic.jpg

Toronto, Canada

I am a Ph.D. student at the University of Toronto, supervised by Prof. Sanja Fidler. I believe that the smarts of our machines are a mirror of our own thoughts, especially when we feed them with carefully curated and labeled datasets and interact with them in meaningful ways. Therefore, my reserach intersts lies in improving dataset labels both in precision and consistency and helping machine to be a better human feedback listener.

Previously, I worked at Vector Institute and USC as a visiting student in 2018 and 2017 respectively blog posts. Before my Ph.D. journey, I studied Electrical Engineering in National Tsing Hua University, supervised by Prof. Min Sun.

news

Jan 15, 2024 :star: Our paper Transferring Labels to Solve Annotation Mismatches Across Object Detection Datasets is accepted to ICLR 2024
Oct 01, 2023 :star: Our paper Bridging the Sim2Real gap with CARE: Supervised Detection Adaptation with Conditional Alignment and Reweighting is accepted to TMLR 2023
Jul 31, 2022 :star: Present “Data Annotation” at NVIDIA Data-centric seminar
Feb 28, 2022 :star: [Feb. 2022 - May 2023] Internship at NVIDIA Toronto, Canada
Dec 01, 2021 :star: One workshop paper accepted to NeurIPS’21 Data-Centric AI workshop and Participate in Graduate Consortium in HCOMP’21 as Graduate Colleagues

selected publications

  1. label_transfer.png
    Translating Labels to Solve Annotation Mismatches Across Object Detection Datasets
    Yuan-Hong Liao, David Acuna , Rafid Mahmood , James Lucas , Viraj Uday Prabhu , and Sanja Fidler
    In The Twelfth International Conference on Learning Representations , 2024
  2. good_practices.png
    Towards Good Practices for Efficiently Annotating Large-Scale Image Classification Datasets
    Yuan-Hong Liao, Amlan Kar , and Sanja Fidler
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , Jun 2021