Research Scientist · Netflix
Zhuoran Yu
I am a Research Scientist at Netflix, working on multimodal models and evaluation for video and visual content. My research interests include multimodal understanding and reasoning, data-centric learning, and benchmarks that reveal model capabilities and limitations.
I received my Ph.D. in Computer Science from UW–Madison, advised by Yong Jae Lee. Previously, I earned my master's in Computer Science at Georgia Tech, where I worked with Zsolt Kira.
Publications
Research & collaborationsNot Another Text Benchmark: Putting the "Visual" Back in Visual Question Answering for Large Video Models
Conference on Neural Information Processing Systems (NeurIPS), 2026, Evaluations & Datasets Track
DocHop: Benchmarking Out-of-domain Multi-hop Reasoning in Information-Dense Documents
International Conference on Machine Learning (ICML), 2026
Large Language Model Teaches Visual Students: Cross-Modality Transfer of Fine-Grained Conceptual Knowledge
International Conference on Machine Learning (ICML), 2026
See, Hear, and Understand: Benchmarking Audiovisual Human Speech Understanding in Multimodal Large Language Models
Findings of IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR Findings), 2026
Revisiting Active Speaker Detection: An In-the-Wild Benchmark for Generalization and Robustness
INTERSPEECH, 2026
Composition-Grounded Instruction Synthesis for Visual Reasoning
International Conference on Learning Representations (ICLR), 2026
DAVE: A VLM Vision Encoder for Document Understanding and Web Agents
International Conference on Learning Representations (ICLR), 2026
How Multimodal LLMs Solve Image Tasks: A Lens on Visual Grounding, Task Reasoning, and Answer Decoding
Conference on Language Modeling (COLM), 2025
LASER: Lip Landmark Assisted Speaker Detection for Robustness
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2026 (Oral)
CuRe: Cultural Gaps in the Long Tail of Text-to-Image Systems
IEEE/CVF International Conference on Computer Vision (ICCV), 2025
Diversify, Don't Fine-Tune: Scaling Up Visual Recognition Training with Synthetic Images
Transactions on Machine Learning Research (TMLR), 2025 (Featured Certificate)
Denoising and Selecting Pseudo-Heatmaps for Semi-Supervised Human Pose Estimation
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2024
InPL: Pseudo-labeling the Inliers First for Imbalanced Semi-supervised Learning
International Conference on Learning Representations (ICLR), 2023
Group R-CNN for Weakly Semi-supervised Object Detection with Points
Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022
Scale-Equalizing Pyramid Convolution for Object Detection
Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020
Experience
- Research Scientist Intern, IBM Watson Lab (May 2025 - August 2025)
- Research Scientist Intern, Meta FAIR (May 2024 - August 2024)
- Research Scientist Intern, Meta CORE AI (May 2023 - August 2024)
- Applied Scientist Intern, AWS Rekognition (May 2022 - August 2022)
- Research Intern, SenseTime (May 2019 - December 2019)
Teaching
- Teaching Assistant: CS 354 Intro to Computer Systems (September 2021 - December 2021)
- Teaching Assistant: CS 7643 Deep Learning (January 2020 - April 2021)
- Teaching Assistant: CS 7641 Machine Learning (January 2019 - April 2019)
Awards
- First-year Scholarship, UW-Madison (2021)
- Arthur Beaumont Memorial Scholarship, University of Waterloo (2018)