Kangjun Liu

Dr. Kangjun Liu

Peng Cheng Laboratory, Shenzhen, Guangdong, China
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 November 2024 in IEEE Transactions on Cognitive Communications and Networking
Verified
Journal Article
Closed access
SZShuhang Zhang
Shuhang Zhang
SJShuai Jiang
Shuai Jiang
WLWanjie Lin
Wanjie Lin
ZFZheng Fang
Zheng Fang
KLKangjun Liu
Kangjun Liu
Ke ChenKe Chen
Ke Chen
Radio map is an efficient demonstration for visually displaying the wireless signal coverage within a certain region. It has been considered to be increasingly helpful for the future sixth generation (6G) of wireless networks, as wireless nodes are becoming more crowded and complicated. However, the construction of high resolution radio map is very challenging due to the sparse sampling in practical systems. Generative artificial intelligence (AI), which is capable to create synthetic data to fill in gaps in real-world measurements, is an effective technique to construct high precision radio maps. Currently, generative models for radio map construction are trained with two-dimension (2D) single band radio maps in urban scenario, which has poor generalization in diverse terrain scenarios, spectrum bands, and heights. To tackle this problem, we provide a multiband three-dimension (3D) radio map dataset with consideration of terrain and climate information, named SpectrumNet. It is the largest radio map dataset in terms of dimensions and scale, which contains the radio map of 3 spacial dimensions, 5 frequency bands, 11 terrain scenarios, and 3 climate scenarios. We introduce the parameters and settings for the SpectrumNet dataset generation, and evaluate four baseline methods for radio map construction based on the SpectrumNet dataset. Experiments show the necessity of the SpectrumNet dataset for training models with strong generalization in spacial, frequency, and terrain scenario domains. Future works on the SpectrumNet dataset are also discussed, including the dataset expansion and calibration, as well as the extended studies on generative models for radio map construction based on the SpectrumNet dataset.

18 August 2022 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
KLKangjun Liu
Kangjun Liu
Ke ChenKe Chen
Ke Chen
KJKui Jia
Kui Jia
Fine-grained visual classification can be addressed by deep representation learning under supervision of manually pre-defined targets ( e.g., one-hot or the Hadamard codes). Such target coding schemes are less flexible to model inter-class correlation and are sensitive to sparse and imbalanced data distribution as well. In light of this, this paper introduces a novel target coding scheme – dynamic target relation graphs (DTRG), which, as an auxiliary feature regularization, is a self-generated structural output to be mapped from input images. Specifically, online computation of class-level feature centers is designed to generate cross-category distance in the representation space, which can thus be depicted by a dynamic graph in a non-parametric manner. Explicitly minimizing intra-class feature variations anchored on those class-level centers can encourage learning of discriminative features. Moreover, owing to exploiting inter-class dependency, the proposed target graphs can alleviate data sparsity and imbalanceness in representation learning. Inspired by recent success of the mixup style data augmentation, this paper introduces randomness into soft construction of dynamic target relation graphs to further explore relation diversity of target classes. Experimental results can demonstrate the effectiveness of our method on a number of diverse benchmarks of multiple visual classification, especially achieving the state-of-the-art performance on three popular fine-grained object benchmarks and superior robustness against sparse and imbalanced data. Source codes are made publicly available at https://github.com/AkonLau/DTRG .

7 November 2022 in Pattern Recognition
Verified
Journal Article
Closed access
ZXZelin Xu
Zelin Xu
KLKangjun Liu
Kangjun Liu
Ke ChenKe Chen
Ke Chen
CDChangxing Ding
Changxing Ding
YWYaowei Wang
Yaowei Wang
KJKui Jia
Kui Jia

26 August 2024 in Pattern Recognition
Verified
Journal Article
Closed access
KLKangjun Liu
Kangjun Liu
Ke ChenKe Chen
Ke Chen
KJKui Jia
Kui Jia
YWYaowei Wang
Yaowei Wang

23 September 2025 in IEEE Transactions on Cognitive Communications and Networking
Verified
Journal Article
Closed access
KLKangjun Liu
Kangjun Liu
CQChunyan Qiu
Chunyan Qiu
Ke ChenKe Chen
Ke Chen
QZQingfang Zheng
Qingfang Zheng
LSLingyang Song
Lingyang Song
YWYaowei Wang
Yaowei Wang
Next-generation communication systems, such as 6G, require real-time, high-resolution modeling of electromagnetic signal propagation to meet the challenges of ultra-high data rates, dynamic environments, and sparse spatial observations. Radio map estimation (RME) plays a critical role in spectrum management by providing spatial visualization of signal distributions. However, existing methods struggle with the complexities of sparse sampling and nonlinear propagation dynamics, often failing to adapt to the demands of dynamic and unpredictable environments. To overcome these challenges, this paper introduces DAT-Unet, a deformable attention transformer-based framework specifically designed for RME under extreme sparsity. DAT-Unet utilizes deformable attention to dynamically adjust receptive fields, enabling it to capture intricate spatial correlations and adapt to complex channel variations. Its geometry-aware modeling integrates spatial priors, while multi-scale contextual fusion enhances feature representation across diverse propagation patterns. Extensive experiments on public datasets demonstrate the superiority of DAT-Unet, achieving state-of-the-art performance in RME tasks, even with spatial sampling rates several orders of magnitude lower than those required by existing methods. By addressing the challenges of sparse sampling and dynamic environmental complexities, this work highlights the potential of deformable attention mechanisms to enhance RME for next-generation communication systems. The code is available at https://github.com/AkonLau/DAT-Unet

8 December 2024
Verified
Conference Paper
Closed access
SSShuai Shao
Shuai Shao
KLKangjun Liu
Kangjun Liu
QLQingyu Liu
Qingyu Liu
SZShuhang Zhang
Shuhang Zhang
KCKe Chen
Ke Chen
LSLingyang Song
Lingyang Song