Working With Me

Thank you for your interest in working together. Below are answers to the questions I am asked most often. Reading them first will help you decide whether we are a good match — and help me give your note the attention it deserves.

What I work on and who is a good fit

I build AI systems for filmmaking and video creation, spanning generative video, multimodal understanding, HCI, graphics, and audio. I work best with people who are genuinely excited about these areas and can point to something they have built or studied that connects to them.

You do not need to have worked on my exact topics, but there should be real overlap between your interests and mine. The notes that stand out show me a specific connection — a project you have done, a paper of mine you engaged with, or a problem you want to tackle — rather than a message sent to many people at once.

Research Interns

Do you host research interns?

Yes — mentoring research interns is one of my favorite parts of the job, and interns play a central role in the systems my team builds. Whether I have an opening in a given cycle depends on my current projects and team needs.

How do I apply?

Two things, ideally both:

Apply through the official Adobe Research internship page, where all intern hiring is processed.
Send me a short email (see How to reach out below) so I know your interests and can look out for your application.

What background helps?

Strong engineering skills and the ability to build nontrivial, end-to-end systems matter most — especially hands-on experience turning generative AI models and APIs into working prototypes. A clear, specific interest in the work matters more than a long CV.

What is an internship on your team like?

Expectations differ across Adobe Research teams, so to be clear about mine: internships here are not measured by publications. We build highly functional systems that showcase the best of generative AI — combining first-party and third-party models — and success looks like a real, working system that demonstrates what is newly possible.

Student Collaborations

I have an idea I would love to work on with you — can we collaborate?

I am always glad to hear about work that connects with mine. If you are a student, the smoothest path is usually to involve your advisor, so the collaboration has support on both sides. A quick note from your advisor introducing the idea goes a long way.

Will you give feedback on my paper or application?

I read what I can, but I am not able to give detailed individual feedback outside of active collaborations and mentoring. Thank you for understanding.

Full-Time Roles at Adobe Research

Do you have full-time openings?

Full-time research positions are posted and handled through Adobe's official channels — I cannot process applications directly. If there is an open role that fits and our interests align, apply through the Adobe Research careers page, and you are welcome to mention your interest in a brief note.

How to reach out

If we have already met or corresponded, or you have a quick question about my research, just email me — the notes below are mainly for people reaching out for the first time. You can reach me at ding@dingzeyu.li.

To help me find and respond to your message:

Add a subject-line tag such as [Internship], [Collaboration], or [Full-time].
Attach your CV.
In two or three sentences, tell me why we are a fit — a paper or project of mine you connect with, and what you would want to work on.
For internships, mention who I may contact as a reference (usually your advisor or a previous mentor).

I read every message, but the volume means I cannot always reply individually — please do not take a quiet inbox personally. If your note is time-sensitive and you have not heard back in a couple of weeks, a gentle follow-up is welcome.

Reviewing and Service

I am glad to contribute to the community. I have served on international program committees for CHI, Eurographics, Graphics Interface, and ACM Multimedia, and as a reviewer for UIST, CHI, SIGGRAPH, CVPR, ICCV, and related venues. I prioritize venues that operate under an open-access model.