Shunlei Li

Dr. Shunlei Li


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
29 January 2026 in Information Fusion
Verified
Journal Article
Closed access
SLShunlei Li
Shunlei Li
LGLongsen Gao
Longsen Gao
JWJin Wang
Jin Wang
CCChang Che
Chang Che
XXXi Xiao
Xi Xiao
Hamid Reza KarimiHamid Reza Karimi
Hamid Reza Karimi

27 March 2024 in IEEE Transactions on Power Electronics
Verified
Journal Article
Closed access
YHYuegang Hu
Yuegang Hu
WZWeishu Zhan
Weishu Zhan
SLShunlei Li
Shunlei Li
MAMuhammad Adeel Azam
Muhammad Adeel Azam
This article introduces a new nonisolated single-switch trans-inverse high step-up semiquadratic dc–dc converter with low input current ripple for renewable energy generation systems. In the presented topology, a three-winding coupled inductor combined with a quadratic boost converter to achieve high voltage gains. The advanced features of the suggested structure are its ultrahigh voltage conversion ratio, low voltage stress ratio across the switching components, low input current ripple, zero current switching of the semiconductors and also common ground between the input and output sides. Due to the trans-inverse feature in the proposed circuit, higher voltage gains can be achieved without needing large turns’ ratios of the coupled inductor in comparison to the other typical quadratic converters, which decreases the conduction power loss. Furthermore, in this topology, the maximum voltage stress across the power switch is mitigated with the help of a regenerative passive clamp cell. The operating principle, steady-state analysis, theoretical efficiency analysis of the suggested circuit, along with performance comparison with other similar converters are discussed in detail. Finally, to verify the theoretical analysis, a 200 W with 25 V input voltage and 400 V output dc voltage sample prototype is established.

25 December 2024 in Engineering Applications of Artificial Intelligence
Verified
Journal Article
Closed access
SLShunlei Li
Shunlei Li
LWLili Wan
Lili Wan
YZYin Zhang
Yin Zhang
LLLixia Luo
Lixia Luo

7 October 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
SLShunlei Li
Shunlei Li
MAMuhammad Adeel Azam
Muhammad Adeel Azam
AGAjay Gunalan
Ajay Gunalan
Leonardo S. MattosLeonardo S. Mattos
Leonardo S. Mattos
Optical coherence tomography (OCT) is a rapidly evolving imaging technology that combines a broadband and low-coherence light source with interferometry and signal processing to produce high-resolution images of living tissues. However, the speckle noise introduced by the low-coherence interferometry and the blur from device motions significantly degrade the quality of OCT images. Convolutional neural networks (CNNs) are a potential solution to deal with these issues and enhance OCT image quality. However, training such networks based on traditional supervised learning methods is impractical due to the lack of clean ground truth images. Consequently, this research proposes an unsupervised learning method for OCT image enhancement, termed one-step enhancer (OSE). Specifically, OSE performs denoising and deblurring based on a single step process. A generative adversarial network (GAN) is used for this. Encoders disentangle the raw images into a content domain, blur domain and noise domain to extract features. Then, the generator can generate clean images from the extracted features. To regularize the distribution range of retrieved blur characteristics, KL divergence loss is employed. Meanwhile, noise patches are enforced to promote more accurate disentanglement. These strategies considerably increase the effectiveness of GAN training for OCT image enhancement when used jointly. Both quantitative and qualitative visual findings demonstrate that the proposed method is effective for OCT image denoising and deblurring. These results are significant not only to provide an enhanced visual experience for clinicians but also to supply good quality data for OCT-guide operations. The enhanced images are needed, e.g., for the development of robust, reliable and accurate autonomous OCT-guided surgical robotic systems.

27 May 2024 in Artificial Intelligence Review
Verified
Journal Article
Open access
PDF file
SLShunlei Li
Shunlei Li
JTJing Tang
Jing Tang
WZWen Zhou
Wen Zhou
ZYZhang Yin
Zhang Yin
MAMuhammad Adeel Azam
Muhammad Adeel Azam
Leonardo S. MattosLeonardo S. Mattos
Leonardo S. Mattos
In recent years, the exponential growth of online social networks as complex networks has presented challenges in expanding networks and forging new connections. Link prediction emerges as a crucial technique to anticipate future relationships among users, leveraging the current network state to address this challenge effectively. While link prediction models on monoplex networks have a well-established history, the exploration of similar tasks on multilayer networks has garnered considerable attention. Extracting topological and multimodal features for weighting links can improve link prediction in weighted complex networks. Meanwhile, establishing reliable and trustworthy paths between users is a useful way to create metrics that convert unweighted to weighted similarity. The local random walk is a widely used technique for predicting links in weighted monoplex networks. The aim of this paper is to develop a semi-local random walk over reliable paths to improve link prediction on a multilayer social network as a complex network, which is denoted as Reliable Multiplex semi-Local Random Walk (RMLRW). RMLRW leverages the semi-local random walk technique over reliable paths, integrating intra-layer and inter-layer information from multiplex features to conduct a trustworthy biased random walk for predicting new links within a target layer of multilayer networks. In order to make RMLRW scalable, we develop a semi-local random walk-based network embedding to represent the network in a lower-dimensional space while preserving its original characteristics. Extensive experimental studies on several real-world multilayer networks demonstrate the performance assurance of RMLRW compared to equivalent methods. Specifically, RMLRW improves the average f-measure of the link prediction by 3.2% and 2.5% compared to SEM-Path and MLRW, respectively.

30 September 2024 in IEEE Transactions on Medical Robotics and Bionics
Verified
Journal Article
Closed access
SLShunlei Li
Shunlei Li
AGAjay Gunalan
Ajay Gunalan
MAMuhammad Adeel Azam
Muhammad Adeel Azam
VPVeronica Penza
Veronica Penza
Darwin G. CaldwellDarwin G. Caldwell
Darwin G. Caldwell
Leonardo S. MattosLeonardo S. Mattos
Leonardo S. Mattos
This paper introduces a new controller for real-time dynamic laser ablation: the autonomous computer-assisted laser microsurgery system (Auto-CALM). Auto-CALM allows the surgeon to define the ablation area, which is then precisely ablated by the system while compensating for tissue motions and deformations. This is achieved based on three control blocks: target tracking, laser tracking, and ablation control algorithm. The ablation area, i.e. the target, is defined by the surgeon using a graphics tablet and graphics overlay on the surgical video. This target is then tracked in real-time using improved optical flow and a novel scaling strategy that makes the system robust against tissue deformations. Laser tracking is based on a pretrained Segment Anything Model that localizes the position of the laser in the surgical video. The ablation algorithm generates a trajectory to ablate the target given the dynamically updated laser position and target position. This enables motion compensation, which increases the accuracy of the system. Auto-CALM was validated through laser ablation experiments based on a porcine larynx fixed to a breathing motion simulation stage. The obtained results were also compared with those achieved under manual operation of CALM, and under autonomous ablation using the Track Anything Model as the target tracking algorithm. Furthermore, four different parts of the ex-vivo porcine larynx were tested to investigate different tracking features and the robustness of the system. Auto-CALM achieved a Dice Similarity Coefficient of 95.49% under the most challenging conditions (including tissue motion and no feature), reaching an ablation speed of 1.43 mm2/s. The accuracy and usability of the integrated platform bear potential for the accurate ablation of tissue volumes in clinical settings. Further ex-vivo and in-vivo animal studies shall help translate these findings to clinical use.