Wenbo LI 李文博 Email: fenglinglwb AT gmail DOT com We are currently seeking highly motivated interns with top-tier publications to join us in the TGT program. The work will focus on unified multimodal models. Please feel free to reach out.

Biography

I am a research scientist at JD Explore Academy, where I focus on multimodal foundation models, AIGC, and low-level vision. Prior to this, I was part of the Huawei TopMinds program (天才少年) at the Huawei Noah's Ark Lab.

I earned my Ph.D. in the CSE Department at the Chinese University of Hong Kong in 2023, under the guidance of Prof. Jiaya Jia. I received my Master's degree in 2019 and Bachelor's degree from the IEEE Pilot Class in 2016 from Shanghai Jiao Tong University.

News

• Four papers accepted at ICCV 2025, including TurboVSR as a Spotlight.
• Two papers accepted at ICML 2025.
• One paper accepted at IJCAI 2025.
• Eight papers accepted at CVPR 2025.
• One paper accepted at ICLR 2025.
• Three papers accepted at AAAI 2025.

Recent Work

• Intelligent multimodal agents: JarvisArt , JarvisIR , and RestoreAgent.
• Efficient image and video synthesis methods: LoViC, Turbo2K, FlashVideo , UltraPixel .
• Efficient controllable image and video generation method: ControlNeXt .
• Efficient diffusion-based image and video SR models: PocketSR, FlowSR, TurboVSR, FluxSR, PassionSR, etc.

Honors and Awards

• Our MAT is selected in the CVPR 2022 Best Paper Finalists
• Champion of MIPI 2025 Challenge on Deblurring for HVS/HybridEVS Camera
• Champion of AIM 2019 Challenge on Video Extreme Super-Resolution
• Champion of COCO 2018 Challenge on Keypoint Detection
• Outstanding Graduate, SJTU
• Academic Excellence Scholarship, SJTU
• Wish Enterprise Scholarship
• Fushou Li Enterprise Scholarship

Academic Service

• TPAMI, IJCV, TIP
• ICML, ICLR, NeurIPS, CVPR, ICCV, ECCV, AAAI, IJCAI