Langezijds

Dr.-Ing. H. Cheng (Hao)

Assistant Professor
ITC-SCI-TECH
Langezijds 1308
h.cheng-2@utwente.nl
+31534892668
Business card

Hao Cheng is an Assistant Professor in the Department of Geoinformatics at the Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, the Netherlands. Previously, he was a researcher and MSCA Postdoctoral Fellow at ITC from 2022 to 2024 and a postdoctoral researcher at the Institute of Cartography and Geoinformatics, Leibniz University Hannover, Germany, from 2021 to 2022. He received his M.Sc. in Internet Technologies and Information Systems in 2017 from TU Braunschweig, Leibniz University Hannover, TU Clausthal, and the University of Göttingen in Germany, and his Ph.D. in 2021 from the Faculty of Civil Engineering and Geodetic Science at Leibniz University Hannover in Germany. He is a member of the ELLIS Society and a co-chair of ISPRS WG I/7, Embodied Sensor Systems and In-Sensor Intelligence for 2026–2030.

Expertise

  • Computer Science

    • Autonomous Driving
    • Deep Learning Method
    • Learning Approach
    • Prediction Model
    • Object Detection
    • Pose Estimation

  • Social Sciences

    • Road User
    • Neural Network

Organisations

His research focuses on accessible and responsible GeoAI and deep learning, particularly perception, reconstruction, and world models.

Publications

2026

DriveVA: Video action models are zero-shot drivers (2026)In Computer Vision - ECCV 2026 (pp. 315-335) (Lecture Notes in Computer Science; Vol. 17066). Springer. Liu, M., Zhang, D., Liu, J., Cui, J., Xie, H., Chen, G., Ye, H., Yang, M. Y., Nex, F. & Cheng, H.https://doi.org/10.1007/978-3-032-37718-0_19AfriTickID: An AI-assisted image-based approach for tick species identification in Kenya. (2026)Acta Tropica, 281. Article 108268. Kioko, C., Cheng, H., Ekeya, J., Wesonga, V., Murigu, M., Githaka, N., Cheng, Y. & Blanford, J. I.https://doi.org/10.1016/j.actatropica.2026.108268Deep learning based polygonal object outline extraction from airborne imagery (2026)[Thesis › PhD Thesis - Research UT, graduation UT]. University of Twente, Faculty of Geo-Information Science and Earth Observation (ITC). Jiao, W.https://doi.org/10.3990/1.9789036573689UNIVERSE: Unified video action models for autonomous driving with flexible mask-modulated modality generation (2026)[Working paper › Preprint]. Liu, M., Zhang, D., Liu, J., Cui, J., Xie, H., Chen, G., Ye, H., Nex, F., Cheng, H. & Yang, M. Y.https://arxiv.org/abs/2607.05133ACPV-Net: All-class polygonal vectorization for seamless vector map generation from aerial imagery (2026)[Contribution to conference › Paper] The IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2026 2026. Jiao, W., Cheng, H., Vosselman, G. & Persello, C.https://openaccess.thecvf.com/content/CVPR2026/papers/Jiao_ACPV-Net_All-Class_Polygonal_Vectorization_for_Seamless_Vector_Map_Generation_from_CVPR_2026_paper.pdfStreamVLO: Streaming visual-LiDAR odometry with cumulative drift compensation (2026)[Contribution to conference › Paper] The IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2026 2026. Liu, M., Liu, J., Yang, M. Y., Jiang, C., Li, J., Zhang, Y., Wang, H., Nex, F. & Cheng, H.https://openaccess.thecvf.com/content/CVPR2026/papers/Liu_StreamVLO_Streaming_Visual-LiDAR_Odometry_with_Cumulative_Drift_Compensation_CVPR_2026_paper.pdfBridge: Basis-driven causal inference marries VFMs for domain generalization (2026)[Working paper › Preprint]. ArXiv.org. Hong, M., Liu, F., Gevaert, C., Vosselman, G. & Cheng, H.https://doi.org/10.48550/arXiv.2604.26820ARFlow: Auto-regressive optical flow estimation for arbitrary-length videos via progressive next-frame forecasting (2026)In International Conference on Learning Representations 2026. Liu, J., Liu, M., Zhu, S., Zhang, Y., Li, J., Yang, M. Y., Nex, F., Cheng, H. & Wang, H.https://proceedings.iclr.cc/paper_files/paper/2026/file/d3408794e41dd23e34634344d662f5e9-Paper-Conference.pdf4DSTR: Advancing Generative 4D Gaussians with Spatial-Temporal Rectification for High-Quality and Consistent 4D Generation (2026)Proceedings of the AAAI Conference on Artificial Intelligence, 40(9), 7224-7232. Liu, M., Liu, J., Zhang, Y., Li, J., Yang, M. Y., Nex, F. & Cheng, H.https://doi.org/10.1609/aaai.v40i9.37659Gaussian on-the-fly splatting: A progressive framework for robust near real-time 3DGS optimization (2026)IEEE Robotics and automation letters, 11(1), 426-433. Xu, Y., Yu, Y., Gan, W., Wang, T., Zhan, Z., Cheng, H. & Wang, X.https://doi.org/10.1109/LRA.2025.3632729

Research profiles

Affiliated study programs

Courses academic year 2026/2027

Courses in the current academic year are added at the moment they are finalised in the Osiris system. Therefore it is possible that the list is not yet complete for the whole academic year.

Courses academic year 2025/2026

HORIZON-TMA-MSCA-PF-EF VeVuSafety-101062870: Traffic safety is a fundamental requirement in vehicular environments and for many artificial intelligence-based systems, such as autonomous vehicles. In urban environments, there are high-risk locations, such as intersections and shared spaces, where vehicles and vulnerable road users (VRUs) interact directly. By advancing state-of-the-art artificial intelligence methodologies, this project, VeVuSafety, aims to develop an end-to-end deep learning framework for learning road users’ behavior in various mixed-traffic scenarios, with the goal of improving the safety of both vehicles and VRUs.

Current projects

VeVuSafety

Traffic safety is the fundamental criterion for vehicular environments and many artificial intelligence-based systems like autonomous vehicles. There are places, e.g., intersections and shared spaces, in urban environments with high risks where vehicles and VRUs directly interact with each other. By advancing state-of-the-art artificial intelligence methodologies, this project VeVuSafety aims to build an end-to-end deep learning framework to learn road users’ behavior in various mixed traffic situations for the safety of vehicles and VRUs.

Address

University of Twente

Langezijds (building no. 19), room 1308
Hallenweg 8
7522 NH Enschede
Netherlands

Navigate to location

Organisations

QR codeScan the QR code or
Download vCard