Senior researcher | Professional Committee @ Chinese Association for Artificial Intelligence (CAAI) Email: jiaojiao.ye@cs.ox.ac.uk [OX Official] | [Profile] | [Wechat]

Now a PhD at the University of Oxford, obtained my Bachelor’s from Zhejiang University.

🚀 Research interest

My research wish is to build a robust, fully automated system 🤖 with learning-based methods in the challenging and uncertain real-world. Particularly, I am keen to conduct research in the interdisciplinary area of Embodied Agents and Artificial Intelligence, with a passion for precise robotic perception and decision-making in a 3D complicated and non-stationary environment.

My research interest is located in the following areas:

• Generative Model
• Embodied AI
• World Model Learning

My Ph.D. work builds on three closely-related questions:

1. How can we efficiently acquire massive, high-quality datasets for scaling foundation models? I explore a training-free generation pipeline that utilizes pre-trained diffusion-based generative models alongside VLMs to synthesize large-scale, well-annotated datasets without incurring additional training costs.
2. How can we enhance a model’s reasoning, planning, and memory capabilities to ensure robust performance in dynamic environments, particularly when addressing complex, long-horizon tasks?
3. How can we optimize a robotic system’s computational efficiency and reduce its processing burden to boost real-time execution in the physical world?

We are looking for Intern with interest on World Model, VLA, Whole-Body Control with Force. Also, I love collaborating with others on research! Feel free to contact me (jiaojiao.ye@cs.ox.ac.uk with topic “Intern/Collaboration + Name” ) if you’re interested in research collaboration, Don’t hesitate! 👏

🆕 News

See my personal news📰 on Linkedin: @JiaojiaoYe

• Paper accepted by [ICIC 2026]
• Paper accepted by [ICASSP 2026]
• Paper accepted by WACV 2025.
• Won ELISA Mobility Grant.
• Started Journey at Oxford.
• Paper accepted by [RA-L Conference].
• Received 1st Scholarships for International Students(5%).
• Started as an Intern at BCAI.

Projects

• Embodied Intelligence: Our team brings deep, multi-disciplinary mastery across the entire robotics spectrum—from high-precision single-arm systems and synchronized dual-arm configurations to advanced, AI-driven humanoids—empowering us to architect agile, resilient solutions for your most complex automation challenges.
• Oneflow: OneFlow is a deep learning framework designed to be user-friendly, scalable and efficient. Since its release, Oneflow has gained nearly 9400 stars GitHub stars and over 1000 forks , users from 200+ top universities & companies(including tsinghua university), demonstrating strong community adoption.

Hornor and Services

• Invited presentation at Oxbridge Conference (Coorganized by Oxford and Cambridge).
• Talk about Sora @ Datawhale, ModelScope [video] [poster]
• Talk about Overview of Stable Diffusion at Datawhale, Epubit, Hugging Face [video] [poster]
• Attended China National Computer Congress by CCF poster
• Tutorial on Inference Speedup tutorial
• Datawhale Contributor.

Teaching

• Artificial Intelligence at the University of Oxford
• Computer Vision at the University of Oxford