Papers
3
Total Citations
38
H-Index
3
About
Victoria Dean is a leading roboticist whose work bridges the gap between simulation and real-world deployment. Her primary research areas include robot learning, manipulation, and multi-modal perception. Dean’s most impactful contribution is the "Train Offline, Test Online" benchmark (2023, 16 citations), which directly tackles three critical bottlenecks in robotics: the high cost of hardware, the lack of standardized platforms, and the scarcity of internet-scale training data. This benchmark provides a reproducible framework for evaluating policies on a real robot after offline training, enabling broader participation and more generalizable findings. In parallel, Dean has advanced multi-modal learning with "Hearing Touch" (2024, 9 citations), demonstrating that audio-visual pretraining can significantly improve contact-rich manipulation tasks—a novel approach that moves beyond vision-only representations. Demonstrating a commitment to responsible innovation, she also co-authored "Teaching Ethics by Teaching Ethics Pedagogy" (2022, 13 citations), which redesigned an Ethics and Robotics course to embed ethical reasoning into the computer science curriculum. With over 38 citations on her core works, Dean is shaping both the technical and ethical foundations of modern robotics.
Research Focus
Key Achievements
Top Papers
- 1Train Offline, Test Online: A Real Robot Learning Benchmark16 citations · 2023
- 2Teaching Ethics by Teaching Ethics Pedagogy13 citations · 2022
- 3Hearing Touch: Audio-Visual Pretraining for Contact-Rich Manipulation9 citations · 2024