Gao Lili
My research focuses on controllable generative modeling for autonomous driving, particularly diffusion-based scene generation to synthesize realistic and rare driving scenarios. I previously worked on synthetic data augmentation approaches to enrich training distributions for downstream tasks such as object detection, as well as diffusion-based scene synthesis with structured conditioning mechanisms to enable controllable scenario generation. I aim to advance conditioned data generation methods that improve robustness and generalization of downstream perception systems.
📬 Contact
- gao (at) zib.de
- languages
- Chinese, English, and German
🎓 Curriculum vitae
- since 2026
- Researcher at ZIB
🔬 Projects
The project aims at shaping the next generation of highway autonomy for heavy trucks by combining long-range multimodal data, controllable generative models, a closed-loop virtual gym, and vision-language model (VLM) embedded reasoning. Rather than treating autonomy as a pure perception or planning problem, the project aims to advance generative simulation, reinforcement learning and chain-of-thought reasoning and language-guided representation learning as a unified research agenda.
