Dr. -Ing. Kunyu Peng

Visiting Postdoctoral Researcher

About Myself:

I am a visiting researcher at the Institute for Computer Science, Artificial Intelligence and Technology (INSAIT), where I am working on multimodal action understanding and video analysis and hosted by Prof. Luc Van Gool and Prof. Danda Paudel.

I lead the action recognition research group in Karlsruhe Institute of Technology, CVHCI Lab, and work there as postdoctoral researcher. I earned my Ph.D. in Informatics Institute of Karlsruhe Institute of Technology, where I was fortunate to be advised by Prof. Dr. -Ing. Rainer Stiefelhagen. Prior to that, I received my M. Sc. degree from ETIT of Karlsruhe Institute of Technology and B. Sc. degree from Beijing Institute of Technology. I finished three research internships at the Chinese Academy of Science (Beijing), ESME Sudria (Paris), and Bosch (Stuttgart). My research interests lie in video understanding, human action understanding, vision language model, and trustworthy challenges.

Honors and Awards:

ICLR2025 Notable Reviewer Award

CVPR2025 Outstanding Reviewer Award

NeurIPS2024 Top Reviewer Award

ICRA2024 Best Paper Finalist

Science, Technology, and Innovation Scholarship

Academic Services

2023-2024 Associated Editor, IV workshop.

2025 Associated Editor, IV main conference.

2023-2025 Reviewer for CVPR, ECCV, ICCV, IROS, ICRA, ICML, NeurIPS, AAAI, AISTATS, ACMMM, IJCNN, ICLR, TITS, TCSVT, TIP, TIV, Nature Artificial Intelligence Review, Nature Scientific Report, Pattern Recognition, RAL, Neurocomputing, Information Fusion.

Research Projects

2021-now SmartAGE, aiming at deep learning-based activity analysis for elderly people, founded by Carl-Zeiss-Foundation, collaborated with Heidelberg university, Frankfurt university, and Mannheim university.

2024-now SFB 1574 Circular Factory, aiming at deep learning based human action segmentation in a circular factory environment, founded by the German Research Foundation, collaborated with Stuttgart university. I take care of PI work tasks in this project.

2025-now HeiKA-Star PACo, aiming at deep learning-based human cognitive status estimation using video data, founded by German Research Foundation, collaborated with Heidelberg university.

Teaching Experiences

WS24/25 Seminar Computer Vision for Human-Computer Interaction,

WS23/24 Seminar Computer Vision for Human-Computer Interaction,

WS22/23 Seminar Computer Vision for Human-Computer Interaction,

SS25 Lecture Deep Learning for Computer Vision I: Basics,

WS24/25 Lecture Deep Learning for Computer Vision II: Advanced Topics,

WS23/24 Lecture Deep Learning for Computer Vision II: Advanced Topics,

2026

Yebin Yang, Di Wen, Lei Qi, Weitong Kong, Junwei Zheng, Ruiping Liu, Yufan Chen, Chengzhi Wu, Kailun Yang, Yuqian Fu, Danda Pani Paudel, Luc Van Gool, Kunyu PengInterEdit: Navigating Text-Guided Multi-Human 3D Motion EditingIn: European Conference on Computer Vision (ECCV 2026)

Ruiping Liu, Yufan Chen, Yuheng Zhang, Junwei Zheng, Kunyu Peng, Chengzhi Wu, Chenguang Huang, Di Wen, Jiaming Zhang, Kailun Yang, Rainer Stiefelhagen
What if? Emulative Simulation with World Models for Situated Reasoning
In: European Conference on Computer Vision (ECCV 2026)

Kai Luo, Fei Teng, Mengfei Duan, Wanjun Jia, Xu Wang, Hao Shi, Kunyu Peng, Zhiyong Li, Kailun Yang
PS-MOT: Cultivating Instance Awareness from Point Seeds for Multi-Object Tracking
In: European Conference on Computer Vision (ECCV 2026)

Yuanfan Zheng, Kunyu Peng, Xu Zheng, Kailun Yang
Seeing Beyond: Extrapolative Domain Adaptive Panoramic Segmentation
In: The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)

Yuheng Zhang, Mengfei Duan, Kunyu Peng, Yuhang Wang, Ruiping Liu, Fei Teng, Kai Luo, Zhiyong Li, Kailun Yang
ProOOD: Prototype-Guided Out-of-Distribution 3D Occupancy Prediction
In: The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)

Junwei Zheng, Ruize Dai, Ruiping Liu, Zichao Zeng, Yufan Chen, Fangjinhua Wang, Kunyu Peng, Kailun Yang, Jiaming Zhang, Rainer Stiefelhagen

RHO: Robust Holistic OSM-Based Metric Cross-View Geo-Localization

In: The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (CVPR 2026)