Lin Zhao

I am a second-year Ph.D. student in College of Engineering, Northeastern University, supervised by Prof. Yanzhi Wang and Prof. Xue Lin. My research interests include deep learning algorithm, multi-modal AI and generative model. Currently, I am focusing on efficiency both at the model level (generative models) and at the data level (datasets). I got my M.S. and B.S. degrees from Nankai University.

Open to collaboration, curious conversations, or just making new friends — feel free to say hi!

Email / Scholar / Twitter / Linkedin / Wechat

Photo by Xinru Jiang

Publications

	Taming Diffusion for Dataset Distillation with High Representativeness Lin Zhao, Yushu Wu, Xinru Jiang, Jianyang Gu, Yanzhi Wang, Xiaolin Xu, Pu Zhao, Xue Lin International Conference on Machine Learning (ICML), 2025 We propose D3HR, which systematically investigates key issues in current diffusion-based methods and, based on these insights, introduces a new paradigm for diffusion-based dataset distillation. [paper] [code]
	FlashEval: Towards Fast and Accurate Evaluation of Text-to-image Diffusion Generative Models Lin Zhao, Tianchen Zhao, Zinan Lin, Xuefei Ning, Guohao Dai, Huazhong Yang, Yu Wang Computer Vision and Pattern Recognition (CVPR), 2024 We propose FlashEval, which can identify a representative subset to speed up the evaluation of text-to-image Diffusion models (10x). [project] [paper] [code]
	THInImg: Cross-modal Steganography for Presenting Talking Heads in Images Lin Zhao, Hongxuan Li, Xuefei Ning, Xinru Jiang Winter Conference on Applications of Computer Vision (WACV), 2024 We propose THInImg, which can present up to 80 seconds of high quality talking-head video (including audio) in an identity image with 160×160 resolution. [paper]
	Deep Symmetric Network for Underexposed Image Enhancement with Recurrent Attentional Learning Lin Zhao, Shaoping Lu, Tao Chen, Zhenglu Yang, Ariel Shamir International Conference on Computer Vision (ICCV), 2021 We propose an invertible framework to solve both underexposed image enhancement and low-light image enhancement problems in a unified structure. [project] [paper] [code]

Work Experiences

	Research Intern, Creative Vision Group, Snap Inc. Jun 2025 - Presents, Santa Monica, CA Worked on Efficient Text-to-video Diffusion
	Research Assistant, NICS-EFC Group, Tsinghua University Jul 2023 - Dec 2023, Beijing, China Worked on Efficient Diffusion Evaluation Professor: Prof. Yu Wang and Prof. Xuefei Ning
	Research Intern, Media Computing Group, Microsoft Research Asia Jul 2022 - Dec 2022, Beijing, China Worked on Talking-head Generation and Video Compression
	Research Intern, Multimodal Group, Sensetime Mar 2022 - May 2022, Beijing, China Worked on Video Editing

Selected Honors & Awards

Gongneng Scholarship. 2021

Graduate Student Scholarship. 2020

Comprehensive First-class Scholarship (Top 5%). 2018, 2019

Honorable Mention, Mathematical Contest in Modeling. 2019

Misc

I was born and raised in Harbin, China.

My baby cat! She is so cute!!

I especially love little monkeys; they are just too adorable!