Doyup Lee

I am the founder of a stealth startup focused on content creation and AI. Before that, I was a senior member of technical staff at Runway, New York, and a Foundational Contributor of Gen-3 Alpha, Frames, and Gen-4, which are the flagship products of Runway for media generation. I was a team lead at Kakao Brain and worked on visual content generation and representation learning. I received my Ph.D. from Pohang University of Science and Technology (POSTECH), South Korea, advised by Prof. Wook-Shin Han, and conducted independent research during the first four years of my Ph.D. I obtained my bachelor degree in Industrial Management Engineering from POSTECH.

News

Aug 2025

One paper was accepted to the main conference track of EMNLP 2025.

Aug 2025

I serve as an Area Chair (AC) for CVPR 2026.

Apr 2025

We released Gen-4, a new generative model for consistent and controllable media.

Jan 2025

We released Frames, a new image generation model with remarkable stylistic control and visual fidelity.

Sep 2024

One paper was accepted to ECCV 2024.

Jul 2024

One paper was accepted to ICML 2024.

Jun 2024

We released Gen-3 Alpha, a new frontier model of video generation with unprecedented quality.

Jan 2024

I joined Runway as a Senior Member of Technical Staff in Applied Research 🎉 !

Dec 2023

One paper was accepted to NeurIPS 2023.

Jun 2023

Three papers were accepted to CVPR 2023, including one highlight paper.

Feb 2022

Two papers were accepted to NeurIPS 2022.

Jun 2022

Two papers were accepted to CVPR 2022.

Selected Publications

Please refer to my Google Scholar page for full publication list.
(* and † indicate co-first author and corresponding author, respectively.)

ECCV 2024

NVS-Adapter: Plug-and-Play Novel View Synthesis from a Single Image

Yoonwoo Jeong, Jinwoo Lee, Chiheon Kim, Minsu Cho†, Doyup Lee†

3D Generation Novel View Synthesis
ICML 2024

Retrieval-Augmented Score Distillation for Text-to-3D Generation

Junyoung Seo, Susung Hong, Wooseok Jang, Min-Seop Kwak, Hyeonsu Kim, Doyup Lee†, Seungryong Kim†

3D Generation RAG Score Distillation
NeurIPS 2023

Locality-Aware Generalizable Implicit Neural Representation

Doyup Lee*, Chiheon Kim*, Minsu Cho, Wook-Shin Han

Implicit Neural Representation Generative Model Image Generation
CVPR 2023

Generalizable Implicit Neural Representations via Instance Pattern Composers

Highlight (acceptance rate: 2.5%)

Chiheon Kim*, Doyup Lee*, Saehoon Kim, Minsu Cho, Wook-Shin Han

Implicit Neural Representation Generalization Audio, Image, 3D
CVPR 2023

Towards End-to-End Generative Modeling of Long Videos With Memory-Efficient Bidirectional Transformers

Jaehoon Yoo, Semin Kim, Doyup Lee, Chiheon Kim, Seunghoon Hong

Video Generation Transformer
NeurIPS 2022

Draft-and-Revise: Effective Image Generation with Contextual RQ-Transformer

Doyup Lee*, Chiheon Kim*, Saehoon Kim, Minsu Cho, Wook-Shin Han

Image Generation Discrete Diffusion Sampling Strategy
CVPR 2022

Autoregressive Image Generation using Residual Quantization

Doyup Lee*, Chiheon Kim*, Saehoon Kim, Minsu Cho, Wook-Shin Han

Image Generation Autoregressive Transformer

Invited Talks

Jun 2025 Invited talk at Kakao Group. "Powering Human Imagination: Visual Content Generation with Large-Scale Generative Models."
May 2025 Invited talk at KAIST AI. "From Video Generative Models to General World Models."
Dec 2024 Invited talk at NeurIPS'24 Workshop on Video-Language Models. "From Video Generative Models to General World Models."
Jul 2024 Invited talks at Samsung Research, Krafton, UNIST, and POSTECH. "Recent Advances in Large-Scale Generative Models for Visual Contents."
Oct 2023 Invited talk at Korea University. "Large-Scale Generative Models for Visual Contents."
Oct 2023 Invited talks at POSTECH and Korea University. "Exploring Generalizable Implicit Neural Representations: An Introduction."
Dec 2022 Invited talk at Korea Software Congress (KSC). "Draft-and-Revise: Effective Image Generation with Contextual RQ-Transformer."
Jul 2021 Invited talk at CVPR'21 Workshop on Open World Vision. "Open World Vision with Contrastive Regularization."

Doyup Lee

News

Selected Publications

NVS-Adapter: Plug-and-Play Novel View Synthesis from a Single Image

Retrieval-Augmented Score Distillation for Text-to-3D Generation

Locality-Aware Generalizable Implicit Neural Representation

Generalizable Implicit Neural Representations via Instance Pattern Composers

Towards End-to-End Generative Modeling of Long Videos With Memory-Efficient Bidirectional Transformers

Draft-and-Revise: Effective Image Generation with Contextual RQ-Transformer

Autoregressive Image Generation using Residual Quantization

Invited Talks

Get In Touch