Jordan Lin

CS + Math @ UCLA | Research @ Zhou Lab | URSP @ UCLA | Prev Research @ Snap

jordan.jpg

Hello! My name is Kuan Heng (Jordan) Lin, and I am a fourth-year Computer Science undergraduate at UCLA with a minor in Mathematics, working with Professor Bolei Zhou on computer vision and generative AI.

My research interests stem from generative AI to topics such as computer vision, multimodal/3D generation, controllable generation, generative vision for autonomy, and foundation & world models. Particularly, my work revolves around pushing the boundaries and applications of large generative models to allow us to interact with them in a more human way.

I am previously a Research Intern at Snap’s Creative Vision team, where I worked on large-scale video diffusion models (SnapVideo V2) with Willi Menapace, particularly building Snap’s first 2D (Data + Model) Parallel framework and implementing variable-length multimodal training. I am also passionate about CS and ML education: I am part of the Program Development Team at the UCLA Learning Assistant Program, and I am an Advisor and Ex-President of ACM AI @ UCLA.

I also do :camera: photography and :movie_camera: filmmaking.

My name: 林宽恒 (lín kuān héng), hence my legal name Kuan Heng Lin.

News

Dec 18, 2024 I was selected as a finalist for the 2024–25 CRA Outstanding Undergraduate Researcher Award :D I want to thank UCLA CS for nominating me, and Sicheng Mo and Prof. Bolei Zhou for all the mentorship along the way!
Sep 25, 2024 Ctrl-X has just been accepted to NeurIPS 2024! :tada: See y’all at Vancouver :D
Aug 20, 2024 I was selected for the 2024–2025 Undergraduate Research Scholars Program (URSP) at UCLA!
Jun 12, 2024 I just released the paper Ctrl-X: Controlling Structure and Appearance for Text-To-Image Generation Without Guidance, where I am co-first author with the incredible Sicheng Mo :D
Feb 26, 2024 FreeControl has just been accepted to CVPR 2024! :tada:

Selected Publications

2024

  1. ctrl-x.jpg
    Ctrl-X: Controlling Structure and Appearance for Text-To-Image Generation Without Guidance
    Kuan Heng Lin*Sicheng Mo*Ben KlingherFangzhou Mu, and Bolei Zhou
    In Neural Information Processing Systems (NeurIPS), 2024
  2. freecontrol.jpg
    FreeControl: Training-Free Spatial Control of Any Text-to-Image Diffusion Model with Any Condition
    Sicheng Mo*Fangzhou Mu*Kuan Heng Lin, Yanli Liu, Bochen Guan, Yin Li, and Bolei Zhou
    In Computer Vision and Pattern Recognition (CVPR), 2024

Recent Projects & Blogs