VIBE Lab
Vision, Intelligence, Brain, Embodiment
We are VIBE Lab at Huazhong University of Science and Technology (HUST) — advancing the understanding and capabilities of generative intelligence systems, bridging the gap between large-scale models and real-world applications.
About
Dr. Tao Hu (胡涛) leads the lab with a focus on pushing the boundaries of generative AI. He received his Ph.D. from the University of Amsterdam in 2023 (B.S. Wuhan University, M.S. Peking University), where he was supervised by Prof. Cees Snoek (VISlab), followed by postdoctoral research at the renowned CompVis Group (Ludwig Maximilian University of Munich) under Prof. Björn Ommer—the team behind Stable Diffusion. His research spans diffusion models, flow matching, and efficient generative architectures, with publications at top-tier venues including CVPR, ICCV, ECCV, NeurIPS, ICLR, AAAI, and T-PAMI. Notable contributions include Self-Guided Diffusion Models (CVPR 2023), which pioneered self-supervised guidance for diffusion; ZigMa (ECCV 2024), the first DiT-style Mamba-based diffusion model; DepthFM (AAAI 2025 Oral), achieving fast monocular depth estimation via flow matching.
Personal Homepage | Google Scholar
Research Areas
- Generative AI
- Large Vision Models - Foundation models, multimodal learning, efficient training and inference
- Embodied AI - Robot learning, sim-to-real transfer, physical intelligence
- Brain-Computer Interface - Neural signal processing, human-machine interaction
What We Offer
- Supportive and collaborative research environment
- Strong mentorship and international collaboration network
- Access to substantial computing resources (GPUs in industry, with ability to rent large cluster GPU resources when necessary)
- Claude Code API reimbursement
- Opportunities to publish at top venues (CVPR, ICCV, NeurIPS, ICLR, etc.)
We are actively recruiting PhD students, Master students, and undergraduate research assistants to join our lab.
Interested? Email your CV, transcript, and a few sentences about your research interests to taohu620 [at] gmail [dot] com.