Research

My research lies at the broader area of computer vision and machine learning, with a particular focus on data- and computation-efficient deep learning and its applications. In the long term, my goal is to develop intelligent systems that can learn from, adapt to, and interact with the environment, and autonomously perform complex tasks in the real world.

Centered on efficient learning, I develop methods that reduce the data and computation cost of deep models across the whole pipeline: efficient input downsampling [1], efficient multi-scale backbones [2], efficient sampling for generative models [3], and efficient neural representations for images [4]. More recently, I have been working on efficient inference for vision-language models (VLMs) through token pruning.

I am also broadly interested in applying these ideas to a range of vision problems, including medical image analysis [5], [6], open-vocabulary and instance segmentation [7], [8], semantic line and skeleton detection [9], [10], salient object detection [11], and face [12] and palmprint [13] recognition.

Taken together, these threads share a common goal: building models that do more with less — less data, less computation, and less supervision — so that intelligent vision systems can scale to real-world problems where efficiency is not a luxury but a necessity. If you are interested in these directions, feel free to reach out.

References

[1]
K. Zhao, L. Ruan, H. Jiang, X. Zhu, X. Zhang, and D. Zeng, “Beyond Predictive Resampling: Learning Input-Agnostic Downsampling for Efficient Aligned Vision Recognition,” in Proceedings of the AAAI Conference on Artificial Intelligence, 2026, pp. 3–11. doi: https://doi.org/10.1609/aaai.v40i16.38319.
[2]
S.-H. Gao, M.-M. Cheng, K. Zhao, X.-Y. Zhang, M.-H. Yang, and P. Torr, “Res2net: A new multi-scale backbone architecture,” IEEE transactions on pattern analysis and machine intelligence, vol. 43, no. 2, pp. 652–662, 2019.
[3]
K. Zhao, A. L. Y. Hung, K. Pang, H. Zheng, and K. Sung, “MRI Super-Resolution with Partial Diffusion Models,” IEEE Transactions on Medical Imaging, 2024, doi: 10.1109/TMI.2024.3483109.
[4]
K. Pang, K. Zhao, A. L. Y. Hung, H. Zheng, R. Yan, and K. Sung, “NExpR: Neural Explicit Representation for fast arbitrary-scale medical image super-resolution,” Computers in Biology and Medicine, vol. 184, p. 109354, 2025, doi: 10.1016/j.compbiomed.2024.109354.
[5]
K. Zhao et al., “PCa-Mamba: Spatiotemporal State Space Models for Prostate Cancer Detection in Multi-Parametric MRI,” Medical Image Analysis, p. 104033, 2026, doi: https://doi.org/10.1016/j.media.2026.104033.
[6]
K. Zhao, K. Pang, A. L. Hung, H. Zheng, R. Yan, and K. Sung, “A Deep Learning-Based Framework for Highly Accelerated Prostate MR Dispersion Imaging,” Cancers, vol. 16, no. 17, p. 2983, 2024, doi: 10.3390/cancers16172983.
[7]
K. Zhao et al., “Open-Vocabulary Camouflaged Object Segmentation with Cascaded Vision Language Models,” Computational Visual Media, vol. 8, no. 3, pp. 331–368, 2025, doi: 10.26599/CVM.2025.9450512.
[8]
X. Wang, K. Zhao, R. Zhang, S. Ding, Y. Wang, and W. Shen, “Contrastmask: Contrastive learning to segment every thing,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 11604–11613.
[9]
K. Zhao, Q. Han, C.-B. Zhang, J. Xu, and M.-M. Cheng, “Deep hough transform for semantic line detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 9, pp. 4793–4806, 2021, doi: 10.1109/TPAMI.2021.3077129.
[10]
K. Zhao, W. Shen, S. Gao, D. Li, and M.-M. Cheng, “Hi-Fi: Hierarchical Feature Integration for Skeleton Detection,” in Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, International Joint Conferences on Artificial Intelligence Organization, Jul. 2018, pp. 1191–1197. doi: 10.24963/ijcai.2018/166.
[11]
K. Zhao, S. Gao, W. Wang, and M.-M. Cheng, “Optimizing the F-measure for threshold-free salient object detection,” in Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 8849–8857.
[12]
K. Zhao, J. Xu, and M.-M. Cheng, “Regularface: Deep face recognition via exclusive regularization,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 1136–1144.
[13]
K. Zhao et al., “Bézierpalm: A free lunch for palmprint recognition,” in European Conference on Computer Vision, Springer, 2022, pp. 19–36. doi: 10.1007/978-3-031-19778-9_2.

This page is written in MDX and the citations are automatically rendered from publications.bib using the rehype-citation plugin.