Tao Wu(吴涛)
My name is Tao Wu, a first-year PhD at the College of Computing and Data Science (CCDS), NTU, supervised by Prof. Yewen Pu. I received my bachelor’s from Northwestern Polytechnical University and my master’s from NTU; during my master’s with Prof. Jianfei Yang I studied noise in human communication for Embodied Question Answering (EQA), which led to NoisyEQA and AbstainEQA. I am now in the Natural Programming Lab, working on human–agent and human–human communication, embodied AI, and human–robot interaction.
Publication
When Robots Should Say "I Don't Know": Benchmarking Abstention in Embodied Question Answering
CVPR 2026 (Highlight)
Abstention benchmarking in embodied QA—when models should decline to answer versus humans.
NoisyEQA: Benchmarking Embodied Question Answering Against Noisy Queries
IROS 2025 Workshop
Benchmarks embodied QA when human queries are noisy or imperfect, bridging to real-world deployment.
News
- April 2026 — When Robots should say I don’t know is selected as a CVPR highlight paper.
- October 2025 — NoisyEQA is accepted to the IROS 2025 workshop.
