OPENINGS · JOIN OUR RESEARCH GROUP
We welcome curious, independent-minded researchers who want to pursue important questions over time—not simply collect publications. Discover research directions, opportunities, and expectations across career stages.
01 / Research focus
Four interconnected research directions spanning medical imaging, physiological and behavioral intelligence, and advanced foundation models.
Lesion detection and segmentation, disease recognition, quantitative analysis, and risk assessment across multi-center and multimodal imaging, with emphasis on generalization and clinical utility.
Integrating EEG, ECG, eye-tracking, video, and behavioral data for temporal representation learning, multimodal fusion, individual state estimation, and robust behavioral understanding.
Pretraining, parameter-efficient adaptation, reinforcement learning-based post-training, knowledge distillation, complex reasoning, and inference acceleration for reliable, efficient models.
Unified representations and alignment across vision, language, EEG, and other physiological signals, enabling joint pretraining and multi-task understanding, reasoning, and generation.
We encourage work across these themes, but prioritize coherent questions and cumulative progress over short-lived trends.
02 / Opportunities
Each stage has a different role. Choose a track to see the expected background and development goals.
Lead independent research and help grow a sustained research direction.
Explore one important problem deeply and develop a coherent research agenda.
Build strong research or engineering ability through a complete project.
Start with understanding, reproduction, benchmarks, and real questions—not publication pressure.
Select a role below to expand details. Milestones are developmental targets—not rigid hiring or degree requirements.
Development targets are discussed according to appointment duration and research maturity; they are not mechanical publication quotas.
Admissions, eligibility, degree requirements, and funded places follow the relevant university and programme regulations.
Research and research-engineering pathways are equally valued when the work is rigorous and reusable.
A paper is not required and is not promised. Strong performers may be invited to participate in longer-term projects, subject to fit, supervision capacity, and applicable programme rules.
All positions, funded places, salaries, and admissions depend on actual openings and the regulations of the relevant institution or programme.
03 / Our approach
We value strong papers, but we care just as much about how good research is done.
Ask why a method works, where it fails, and what is not yet understood.
Go beyond reproducing numbers; question assumptions and test meaningful hypotheses.
Make experiments reproducible, report negative results, and respect evidence.
Take responsibility, communicate clearly, and follow difficult questions through.
Build reusable methods, datasets, systems, and knowledge with teammates.
04 / Research environment
We aim to provide an environment where good ideas can develop through independent work and interdisciplinary collaboration.
High-performance GPU resources and opportunities to work with contemporary foundation models.
Academic collaboration across computing, AI, medical research, and clinical partners.
Opportunities for international academic exchange and collaborative visits, subject to eligibility and available support.
Mentorship appropriate to career stage, with an emphasis on scientific independence and long-term growth.
05 / Get in touch
A concise, specific email is more useful than a heavily polished generic research statement.
Send a CV and a short description of your contributions, research interests, and the problems you want to explore. Include representative papers or projects if available.
No elaborate proposal is required. Tell us what you have studied or reproduced, what questions you found interesting, and how much time you can commit.