RAIR Lab
Ren AI Research (RAIR) Lab
University of North Carolina at Chapel Hill
Our lab works across a broad range of topics in artificial intelligence as the field continues to grow more unified. We are particularly interested in multimodal generation and understanding, with a focus on how AI systems comprehend and engage with complex real-world environments.
Our research philosophy
Why? Openness accelerates scientific progress, builds trust, and enables others to build on our work. We believe impactful research should be reproducible and accessible.
Why? Chasing paper quantity often leads to superficial contributions. We focus on work that solves real problems, influences the field, or creates tools others rely on.
Why? Big leaps often come from ambitious ideas. We support risk-taking and long-term thinking, even if it takes longer to produce results.
Why? Mastery in a niche allows researchers to make meaningful impact and become thought leaders. Broad knowledge is valuable, but depth is what drives innovation.
Why? A great idea is only impactful if others understand and care about it. Salesmanship helps get your work noticed, adopted, cited, and funded. It's not just about making your work sound good — it's about making its value unmistakable.
🎓 Students
PhD
Information for prospective students and visitors
Postdoc: I don’t have a set plan at the moment, but I’m open to considering requests on a case-by-case basis. As I’m still early in my career, it’s likely that I would co-host you with a more senior faculty member. Feel free to email me directly with your CV.
PhD: I’m looking to recruit 1~2 PhD students to join my lab starting Fall 2027. I’m particularly interested in candidates with a background or strong interest in visual generation, 3D vision, VLMs, and multimodal learning. To ensure your application is considered, please follow below steps:
- Read the common questions below with tags [All] and [PhD].
- Apply to the UNC computer science PhD program (The GRE is not required). Be sure to mention my name in your application. Also, I highly recommend submitting your application before December 16, 2025 to receive full consideration for department fellowships!
- Shortlisted applicants will be invited for interviews between early January and Feburary.
- Admitted students will be invited to visit campus in March, with an opportunity to tour the lab and meet the group.
Master’s and undergrad: We welcome UNC students interested in research opportunities to join our lab. Preference is generally given to students who plan to pursue a PhD after completing their current degree. To get started,
- Read the common questions below with tags [All], [MS], and [Undergrad].
- Master’s students: Please reach out only after you have been admitted to UNC or if you are currently enrolled.
- Undergrad students: Preference is typically given to students who are not in their final year.
- Complete the MS/Undergrad/visitors application form.
- If there is mutual interest, we typically follow up within 2 weeks.
Visitor: We’re happy to host highly motivated visiting students in our lab. A minimum stay of 1 semester or 4 months is generally required. To get started,
- Read the common questions below with tags [All].
- Complete the MS/Undergrad/visitors application form.
- If there is mutual interest, we typically follow up within 2 weeks.
🙋 Common Questions
AllWhat are your expectations?
- Self-motivation and passion are the most important qualities I value. Students should be proactive, curious, and eager to learn and explore independently.
- A strong undergraduate-level foundation in programming, mathematics/statistics, machine learning, or computer vision is required. A deep understanding in one or more of these areas is a significant advantage.
- Teamwork is essential — you won’t be working in isolation. We collaborate closely, so being supportive and helpful to others is highly valued.
- Previous research experience is preferred, though not strictly required.
- We strongly encourage diversity and welcome applicants from all backgrounds.
AllWhat's your advising style?
- I would describe my advising style as hands-on while remaining flexible to accommodate individual needs. I usually meet with students at least once per week. I like 30-minute meetings to stay focused and efficient. In addition, there will be weekly group meeting and reading groups.
- As students progress, I encourage them to become increasingly independent. This includes identifying their own research directions, mentoring junior students, and contributing to the academic community in meaningful ways.
- Students are free to work wherever and whenever they feel most comfortable. I do not impose restrictions on work location or hours. However, regular meetings will be held in person.
AllCan I speak with your current or former students/collaborators?
Of course, feel free to reach out.
AllWork-life balance
Maintaining a healthy work-life balance is important. I do not expect students to work evenings, weekends, or holidays, and I fully respect that everyone has their own schedule and responsibilities outside of research. How you manage your time is ultimately up to you—as long as you're making consistent progress and maintaining clear communication, I'm flexible. That said, if something urgent (eg, deadlines) comes up, I'm always happy to make time to discuss it.
AllConference/Travel
Travel expenses for the first author of an accepted paper are typically covered by default. Support for additional authors or other travel-related requests will depend on available funding. In many cases, the CS department or the university also offers additional travel grants, which students are encouraged to apply for.
AllAuthorship
- In most cases, the first author is the person who contributes the majority of the work—typically around 90% or more. The primary advising faculty member is usually listed as the last author.
- Equal contribution is appropriate when two authors contribute in comparable and significant ways, such as developing separate core components of a system.
- I strongly encourage a collaborative lab environment. The first author is responsible for initiating open communication and reaching agreement on authorship order with all collaborators.
- Regardless of who originally proposed the idea, the student who carries out the majority of the work should be listed as the first author.
MS/UndergradRecommendation letters
- There is no strict limit on the number of letters I can write.
- I typically write letters during Fall break. Feel free to send me reminders if my letters are late.
- I only write letters for students who have worked with me directly. Please note that strong recommendation letters are often comparative in nature—so the more experience I have working with you, the more detailed and compelling your letter is likely to be.
MS/PhDInternship
I encourage my students to pursue research internships with industry collaborators, starting as early as the 1st summer. For international students, please be mindful of visa regulations — in particular, the 12-month limit on full-time CPT.
MS/PhDCo-advising
I am open to co-advising students with other faculty members, especially when it adds value to the student's research and overall development. That said, it's important to have a clear discussion in advance about funding responsibilities and mutual expectations. If you're considering a co-advising arrangement, please bring it up as early as possible so we can plan accordingly.
PhDTimeline/graduation criteria
There is no fixed rule for graduation. Decisions are made on a case-by-case basis, typically when a student has developed a strong research portfolio. This is often reflected in multiple high-quality publications and the ability to secure an excellent job offer in either industry or academia. My primary goal is to support each student in reaching a successful and meaningful next step in their career.
PhDWhich fellowship should I apply and when?
"All" and "the earlier the better".
PhDRA & TA
I aim to support my PhD students with research assistantships (RA) throughout their entire program, as long as funding allows. Each student is expected to serve as a teaching assistant (TA) at least once, typically in an AI-related course. Any additional teaching assignments beyond that are optional and based on the student's interests.