Embodied intelligence
Vision-language-action models and robot learning: connecting observations and instructions to useful actions.
PhD researcher · University of Exeter
Robotics &
Embodied AI.
I'm Quangao Liu. I explore how robots can connect perception, reasoning, and action in the physical world.
My interests span vision-language-action models, multimodal perception, and deploying learning systems on real robots.
01 / Research
From perception to action.
Multimodal perception
RGB-D, thermal, LiDAR, and event cameras, with an interest in active perception and complementary sensor information.
Real-world robotics
Sensor integration and practical deployment, exploring what it takes to make learning systems work on physical robots.
02 / Selected projects
Building, learning, testing.
Robot learning / Simulation
3jsVLA
A compact, browser-based environment for learning the vision-language-action pipeline.
Simulate robotic manipulation, collect demonstrations, and train lightweight policies with camera observations, language instructions, and robot state.
Explore the projectTime series / Machine learning
MSTVI
Multi-Scale Time-Variable Interaction for Multivariate Time Series Forecasting.
Research code accompanying our 2025 paper in Knowledge-Based Systems.
03 / Selected publications
Published research.
Prompt2Act: Mapping prompts into sequence of robotic actions with large foundation models
Information Fusion, 127, Part C, 103923 (2026).
TAPO: Dynamic Teacher and Perturbed Answer Injection for Policy Optimization
AAAI 2026, Special Track on AI Alignment, 40(44), 37462–37471.
DAAC: Discrepancy-Aware Adaptive Contrastive Learning for Medical Time series
Advances in Neural Information Processing Systems 38 (NeurIPS 2025), Main Conference Track.
MSTVI: Multi-Scale Time-Variable Interaction for Multivariate Time Series Forecasting
Knowledge-Based Systems, 319, 113551.
FAITH: Frequency-domain Attention In Two Horizons for time series forecasting
Knowledge-Based Systems, 309, 112790.
04 / Get in touch
Let's build something useful.
Interested in robotics, embodied AI, or a research collaboration?
I'd be happy to connect.