Profile image
Kai Wang
Research Fellow
School of Computing
National University of Singapore

About Me

I am a research fellow at the HPC-AI Lab, National University of Singapore, working with Prof. Yang YOU. I also collaborate closely with Prof. Atlas Wang and Prof.Konstantinos N. Plataniotis.

My doctoral thesis committee is composed of Professor Tat Seng CHUA, who serves as the Chair, as well as Professor Mike SHOU and Professor Yang YOU.

I received my M.S. degree in 2020 from MMLab@SIAT , supervised by Prof. Yu Qiao and Assoc. Prof. Xiaojiang Peng .

My research focuses on Parameter Generation and Efficient Machine Learning. I place an emphasis on exploring simple and baseline approaches to gain insights into the workings of deep learning. Until now, my top-5 publications are RPG (arXiv 2025), p-diff (arXiv 2024), PAB (ICLR-2025), CAFE (CVPR-2022), and InfoBatch (ICLR-2024 Oral).

💥 News: I am looking for talented individuals to join my research on Parameter Generation (work with professors and researchers from top universities and institutions). If you are interested, feel free to reach out!

News

  • Feb. 2025: Four papers have been accepted by CVPR-2025. Congrats to Zekai, Wangbo, Mingjia, Zhenghao, Haoxuan, and Yukun!
  • Feb. 2025: Pyramid Attention Broadcast was merged into HuggingFace Diffuers. Stay tuned!
  • Jan. 2025: Pass the thesis defense!
  • Dec. 2024: Enhance-A-Video was merged into ComfyUI-Cogvideo, ComfyUI-LTX, and ComfyUI-Hunyuan. Many thanks to kijai 👏! Stay tuned!
  • Sep. 2024: Pyramid Attention Broadcast was cited by Meta Movie Gen Report. Stay tuned!
  • Sep. 2024: Congrats to my coauthors for their papers accepted to NeurIPS-2024!
  • July 2024: Dataset Growth has been accepted by ECCV-2024. Congrats to my coauthors!
  • May. 2024: Three papers have been accepted by ICML-2024. Congrats to my coauthors!
  • Jan. 2024:InfoBatch, the first lossless dynamic data pruning approach, was selected as Oral Presentation in ICLR-2024..
  • Jan. 2024:DATM, the first lossless dataset distillation approach, was accepted by ICLR-2024..
  • Jan. 2024: Four papers have been accepted by ICLR-2024, including One Oral, One Spotlight, and 2 Posters
  • Sep. 2023: SGDD, GIF, and pFedBreD have been accepted by NeurIPS-2023.
  • July 2023: Dream, Dataset Quantization, and ZSCL have been accepted by ICCV-2023.
  • Feb. 2023: BiCro and MSINet have been accepted by CVPR-2023, two papers consider the data issues in AI.
  • Jan. 2023: Beta was selected as Spotlight Paper in ICLR-2023.
  • Nov. 2022: HaBa was selected as Spotlight Paper in NeurIPS-2022.
  • Sep. 2022: One Paper has been accepted by NeurIPS-2022.
  • July 2022: One Paper has been accepted by ECCV-2022.
  • Mar. 2022: ContrastiveCrop was selected as Oral Presentation in CVPR-2022.
  • Mar. 2022: Give a talk in Sun Yat-sen University, invited by Prof. Xiaodan Liang. Slides coming soon.
  • Mar. 2022: Give a talk in ByteDance, invited by Prof. Jiashi Feng. Slides coming soon.
  • Mar. 2022: Five papers have been accepted by CVPR-2022, including two first-author papers, one equal-first-author paper, and one second-author paper.
  • Oct. 2021: Our Team win one Champion and and one Third Place in ICCV-MFR Challenge, one Second Place in InsightFace Track.
  • Aug. 2021: One Conference paper has been accepted by ICCV-MFR-Workshop.
  • May 2021: Got AI Singapore Ph.D. Fellowship!!!!
  • May 2021: One Journal paper has been accepted by Transaction on Artificial Intelligence.
  • Mar. 2021: Got a NUS Ph.D offer!!!
  • Feb. 2021: One Journal paper has been accepted by Neurocomputing.
  • Jul. 2020: Start my full-time intern at Alibaba DAMO Acadmey.
  • Jul. 2020: One Conference paper has been accepted by ECCV2020.
  • May. 2020: One Conference paper has been accepted by ICIP2020.
  • Feb. 2020: One Conference paper has been accepted by CVPR2020.
  • Jan. 2020: Start my part-time intern at Huawei Cloud.
  • Dec. 2019: One Journal paper has been accepted by TIP.
  • Dec. 2019: One Conference paper has been accepted by AAAI2020.
  • Dec. 2019: Our team achieve Top 6/2612 in NAIC Challenge.
  • Oct. 2019: Three Conference papers have been accepted by ICMI-2019.
  • Sep. 2019: Our team win the Champion in Huawei Garbege Classification Challenge.
  • Aug. 2019: Our team win the Two Second and One Third Place in EmotiW-2019 Challenge.
  • Jan. 2019: Our team win the Champion in PEER (Hosted by UC. Berkeley) Challenge.
  • Aug. 2018: Our team win the One Champion and One Second Place in EmotiW-2018 Challenge.
  • Aug. 2017: Our team win the One Champion in EmotiW-2017 Challenge.
  • Some Awards before 2017, you can check out from my CV

Selected Publications

Scaling Up Parameter Generation: A Recurrent Diffusion Approach
Kai Wang*, Dongwen Tang*, Wangbo Zhao, Konstantin Schürholt, Zhangyang Wang+, and Yang You+
tech report 2025
Conference
Paper Code Twitter


Enhance-A-Video: Better Generated Video for Free
Yang Luo, Xuanlei Zhao, Mengzhao Chen, Kaipeng Zhang, Wenqi Shao+, Kai Wang+, Zhangyang Wang, and Yang You (Corresponding authors)
tech report 2024
Conference
Paper Code Twitter


Emphasizing Discriminative Features for Dataset Distillation in Complex Scenarios
Kai Wang*, Zekai Li*, Zhi-Qi Cheng, Samir Khaki, Ahmad Sajedi, Ramakrishna Vedantam, Konstantinos N Plataniotis, Alexander Hauptmann, and Yang You (Equal contribution)
CVPR 2025
Conference
Paper Code Summary


Real-Time Video Generation with Pyramid Attention Broadcast
Xuanlei Zhao*, Xiaolong Jin*, Kai Wang*+, and Yang You+ (Equal contribution and equal advising)
ICLR-2025
Conference
Paper Code Twitter


A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training
Kai Wang*, Mingjia Shi*, Zekai Li, Yukun Zhou, Zhihang Yuan, Yuzhang Shang, Xiaojiang Peng, Hanwang Zhang, and Yang You (Equal contribution)
CVPR 2025
Conference
Paper Code Summary


Neural Network Diffusion
Kai Wang, Dongwen Tang, Boya Zeng, Yida Yin, Zhaopan Xu, Yukun Zhou, Zelin Zang, Trevor Darrell, Zhuang Liu*, and Yang You* (Equal advising)
arXiv 2024
Conference
Paper Code Twitter


InfoBatch: Lossless Training Speed Up by Unbiased Dynamic Data Pruning
Ziheng Qin, Kai Wang* , Zangwei Zheng, Jianyang Gu, Xiangyu Peng, Zhaopan Xu, Daquan Zhou, Lei Shang, Baigui Sun, Xuansong Xie, and Yang You (Project lead and Equal-first author)
Oral Presentation (1.2%) in The Twelfth International Conference on Learning Representations (ICLR) 2024
Conference
Paper Code OpenReview


Towards Lossless Dataset Distillation via Difficulty-Aligned Trajectory Matching
Ziyao Guo, Kai Wang* , George Cazenavette, Hui Li, Kaipeng Zhang, and Yang You (Project lead )
The Twelfth International Conference on Learning Representations (ICLR) 2024
Conference
Paper Code OpenReview

Dataset Quantization
Daquan Zhou, Kai Wang* , Jianyang Gu*, Xiangyu Peng, Dongze Lian, Yifan Zhang, Yang You, and Jiashi Feng (Equal-first author)
International Conference on Computer Vision (ICCV), 2023
Conference
Paper Code


DREAM: Efficient Dataset Distillation by Representative Matching
Yanqing Liu, Jianyang Gu, Kai Wang* , Zheng Zhu, Wei Jiang, and Yang You (Project Lead)
International Conference on Computer Vision (ICCV), 2023
Conference
Paper Code


DiM: Distilling Dataset into Generative Model
Kai Wang, Jianyang Gu, Daquan Zhou, Zheng Zhu, Wei Jiang, and Yang You
arXiv 2023
Conference
Paper Code


CAFE: Learning to Condense Dataset by Aligning Features
Kai Wang, Bo Zhao，Xiangyu Peng, Zheng Zhu, Shuo Yang, Shuo Wang, Guan Huang, Hakan Bilen, Xinchao Wang, and Yang You
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022
Conference
Paper Code


An Efficient Training Approach for Very Large Scale Face Recognition
Kai Wang, Shuo Wang, Panpan Zhang, Zhipeng Zhou, Zheng Zhu, Xiaobo Wang, Xiaojiang Peng, Baigui Sun, Hao Li, and Yang You
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022
Conference
Paper Code


Crafting Better Contrastive Views for Siamese Representation Learning
Xiangyu Peng, Kai Wang* , Zheng Zhu, Mang Wang, and Yang You (Equal-first author)
Oral Presentation in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022
Conference
Paper Code


Suppressing uncertainties for large-scale facial expression recognition
Kai Wang, Xiaojiang Peng, Jianfei Yang, Shijian Lu, and Yu Qiao
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020
Conference
Paper Code


Region attention networks for pose and occlusion robust facial expression recognition
Kai Wang, Xiaojiang Peng, Jianfei Yang, Debin Meng, and Yu Qiao
IEEE Transaction on Image Processing (TIP), 2020
Journal
Paper Code


Education

Ph.D. in Data Science & Machine Learning
National University of Singapore, Singapore
Aug. 2021 - Present
M.S. in Computer Technology
University of Chinese Acadmey of Sciences
Sep. 2017 - Aug. 2020
B.Eng. in Electrical Engineering and Automation
Beijing Normal University, Zhuhai, China
Sep. 2013 - July 2017

Advising Experience

Current Research Interns

  • ZeKai Li (undergraduate student at NUS, CVPR-2025).
  • Dongwen Tang (undergraduate student at Xidian University).
  • Zhiyuan Liang (undergraduate student at USTC, ICLR-2025).

Alumni

  • Yukun Zhou (undergraduate student at Xidian University, CVPR-2025).
  • Tianyi Li (Ph.D. student at HKU).
  • Ziyao Guo (Ph.D. student at NUS, ICLR-2024, ICML-2024).
  • Yuchen Zhang (incoming Ph.D. student under PKU and Shanghai AI Lab, ICML-2024).
  • Tianle Zhang (Ph.D. student at Nanyang Tech. Univ., Singapore, ICML-2024).
  • Mingjia Shi (incoming Ph.D. student at The University of Virginia, NeruIPS-2023, CVPR-2025).
  • Yanqing Liu (Ph.D. student at UCSC, ICCV-2023).
  • Beining Yang (Ph.D. student at The University of Edinburgh, NeruIPS-2023).

(All talents are welcome to send an email to me if you are interested in collaborating on projects related to Data-Centric AI, Parameter Generation, or other promising research directions.)

Academic Service

  • Reviewer of CVPR, ECCV, ICML, NeurIPS, ICCV, AAAI, ICLR
  • Reviewer of TIP, TCSVT, TAFFEC, TNNLS

    • Friends

    Jianfei Yang (NTU), Huaizheng Zhang (NTU), Konstantinos N Plataniotis (University of Toronto), Zhuang Liu (Meta, FAIR), Atlas Wang (Ut Austin), Yan Yan (University of Illinois Chicago), Chenliang Xu (University of Rochester), Yapeng Tian (UT Dallas), Ramakrishna Vedantam (Self Employed; Prev: Meta AI, FAIR), Zhijian Liu (UCSD), Tianlong Chen (UNC Chapel Hill), Pan He (Auburn University), Xindi Wu (Princeton), George Cazenavette (MIT CSAIL), Konstantin Schürholt (University of St.Gallen), Sijie Ji (NTU), Zheng Zhu (Tsinghua University), Bo Zhao (The University of Edinburgh), Jiankang Deng (Huawei London), Shuyang Sun (Oxford University), Kaipeng Zhang (The University of Tokyo), Hao Luo (Alibaba Group), Xiangyu Peng (NUS), Wangbo Zhao (NUS), Ruicheng Feng (NTU), David Junhao Zhang (NUS), Zangwei Zheng (NUS), Wangbo Zhao (NUS), Yong Liu (NUS), Yaqi Xie (CMU), Fuzhao Xue (Google DeepMind), Kun Wang (NTU)

    Copyright © Kai Wang 2021

    30,285 Total Pageviews