Qinghua Lin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Qinghua Lin has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Zuoyong Li
Minjiang University
7 shared publications - Haoyi Fan
School of Computer and Artificial Intelligence, Zhengzhou University, Zhengzhou 450001, China
3 shared publications - Kun Zeng
Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, School of Computer and Big Data, Minjiang University, Fuzhou, China
3 shared publications - David Zhang
The Chinese University of Hong Kong
2 shared publications - Jiawei Wu
Sun Yat-sen University, Shenzhen Campus
2 shared publications - Rong Hu
School of Computer Science and Mathematics, Fujian University of Technology, Fuzhou, China
2 shared publications
Latest Publications
- 25 March 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 111, 104055
Ultrasound Localization Microscopy Learned from power doppler by uncertainty frequency density estimation and semantic consistency awareness
- 25 March 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 111, 104055
Ultrasound Localization Microscopy Learned from power doppler by uncertainty frequency density estimation and semantic consistency awareness
- Open Access
- 18 March 2026
- Published by Elsevier in SSRN Electronic Journal
Residual Pseudo-Anomaly & Frequency-Domain TripletContrastive Learning for Unsupervised MRI-BasedProstate Cancer Detection
Detecting cancerous anomalies in prostate MRI plays a vital role in the earlydiagnosis of prostate cancer. However, accurate identification remains highlychallenging due to the small size and heterogeneous morphology of cancer-ous regions, as well as the scarcity of fine-grained lesion annotations. To ad-dress these issues, we propose a Pseudo-Anomaly Generation and Frequency-Enhanced Network (PAF-Net) for prostate cancer detection. Specifically, wefirst train an ensemble of autoencoders and constrain their latent featurespaces to encourage diverse and redundancy-reduced representations of nor-mal anatomy. Then, guided by the reconstruction residual discrepanciesamong different autoencoders, we introduce residual-guided perturbationsin anatomically plausible regions to synthesize pseudo-anomaly samples as hard non-normal examples, thereby enriching the decision boundary withoutrequiring manual lesion annotations. Finally, we adopt a hybrid ConvNeXt–Swin detector as the backbone and further incorporate a frequency-domaincontrastive triplet learning strategy on the transformed features to enhancethe discriminability between normal and pseudo-anomalous patterns. Ex-perimental results on the PICAI, Prostate MRI-US, and Brain Tumor MRIdatasets demonstrate that our method achieves AUCs of 85.63%, 90.73%,and 90.83%, and APs of 82.74%, 88.26%, and 88.80%, respectively, outper-forming existing state-of-the-art methods and validating its effectiveness androbustness in prostate and brain MRI anomaly detection tasks.
- 18 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessResidual Pseudo-Anomaly & Frequency-Domain TripletContrastive Learning for Unsupervised MRI-BasedProstate Cancer Detection
Detecting cancerous anomalies in prostate MRI plays a vital role in the earlydiagnosis of prostate cancer. However, accurate identification remains highlychallenging due to the small size and heterogeneous morphology of cancer-ous regions, as well as the scarcity of fine-grained lesion annotations. To ad-dress these issues, we propose a Pseudo-Anomaly Generation and Frequency-Enhanced Network (PAF-Net) for prostate cancer detection. Specifically, wefirst train an ensemble of autoencoders and constrain their latent featurespaces to encourage diverse and redundancy-reduced representations of nor-mal anatomy. Then, guided by the reconstruction residual discrepanciesamong different autoencoders, we introduce residual-guided perturbationsin anatomically plausible regions to synthesize pseudo-anomaly samples as hard non-normal examples, thereby enriching the decision boundary withoutrequiring manual lesion annotations. Finally, we adopt a hybrid ConvNeXt–Swin detector as the backbone and further incorporate a frequency-domaincontrastive triplet learning strategy on the transformed features to enhancethe discriminability between normal and pseudo-anomalous patterns. Ex-perimental results on the PICAI, Prostate MRI-US, and Brain Tumor MRIdatasets demonstrate that our method achieves AUCs of 85.63%, 90.73%,and 90.83%, and APs of 82.74%, 88.26%, and 88.80%, respectively, outper-forming existing state-of-the-art methods and validating its effectiveness androbustness in prostate and brain MRI anomaly detection tasks.
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 4728-4737
Gradient-Aware Revitalization of Non-Effective Samples in Medical Image Segmentation
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 4728-4737
Open AccessGradient-Aware Revitalization of Non-Effective Samples in Medical Image Segmentation
- 25 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
KAC-Unet: A Medical Image Segmentation With the Adaptive Group Strategy and Kolmogorov-Arnold Network
In the field of deep learning-based medical image segmentation, convolutional neural networks extract image features by combining linear convolutional layers with nonlinear activation functions. However, excessive stacking of linear layers in the network limits the model’s ability to capture fine-grained details. In addition, the feature distribution imbalance caused by the traditional fixed grouping strategy can affect the deep model’s capacity to perceive the overall structure of the image. To address these challenges, we propose a medical image segmentation framework, called KAC-Unet. Firstly, we propose the Adaptive Group Strategy to balance the grouping of different input channels, alleviating the performance degradation caused by differences in group information. Then, we propose the Shift Tokenized Kolmogorov-Arnold Network Block to capture complex features in medical images through flexible nonlinear transformations and shift operations. Extensive experiments are conducted on three medical image segmentation datasets. The results demonstrate the effectiveness and superiority of our proposed method compared with state-of-the-art algorithms.
- 25 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
KAC-Unet: A Medical Image Segmentation With the Adaptive Group Strategy and Kolmogorov-Arnold Network
In the field of deep learning-based medical image segmentation, convolutional neural networks extract image features by combining linear convolutional layers with nonlinear activation functions. However, excessive stacking of linear layers in the network limits the model’s ability to capture fine-grained details. In addition, the feature distribution imbalance caused by the traditional fixed grouping strategy can affect the deep model’s capacity to perceive the overall structure of the image. To address these challenges, we propose a medical image segmentation framework, called KAC-Unet. Firstly, we propose the Adaptive Group Strategy to balance the grouping of different input channels, alleviating the performance degradation caused by differences in group information. Then, we propose the Shift Tokenized Kolmogorov-Arnold Network Block to capture complex features in medical images through flexible nonlinear transformations and shift operations. Extensive experiments are conducted on three medical image segmentation datasets. The results demonstrate the effectiveness and superiority of our proposed method compared with state-of-the-art algorithms.
- 11 February 2025
- Published by Elsevier in Pattern Recognition
- Vol. 162
WtNGAN: Unpaired image translation from white light images to narrow-band images
- 11 February 2025
- Published by Elsevier in Pattern Recognition
- Vol. 162
WtNGAN: Unpaired image translation from white light images to narrow-band images
- 4 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (3) , 2603-2615
SIAVC: Semi-Supervised Framework for Industrial Accident Video Classification
Semi-supervised learning suffers from the imbalance of labeled and unlabeled training data in the video surveillance scenario. In this paper, we propose a new semi-supervised learning method called SIAVC for industrial accident video classification. Specifically, we design a video augmentation module called the Super Augmentation Block (SAB). SAB adds Gaussian noise and randomly masks video frames according to historical loss on the unlabeled data for model optimization. Then, we propose a Video Cross-set Augmentation Module (VCAM) to generate diverse pseudo-label samples from the high-confidence unlabeled samples, which alleviates the mismatch of sampling experience and provides high-quality training data. Additionally, we construct a new industrial accident surveillance video dataset with frame-level annotation, namely ECA9, to evaluate our proposed method. Compared with the state-of-the-art semi-supervised learning based methods, SIAVC demonstrates outstanding video classification performance, achieving 88.76% and 89.13% accuracy on ECA9 and Fire Detection datasets, respectively. The source code and the constructed dataset ECA9 will be released in https://github.com/AlchemyEmperor/SIAVC.
- 4 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (3) , 2603-2615
SIAVC: Semi-Supervised Framework for Industrial Accident Video Classification
Semi-supervised learning suffers from the imbalance of labeled and unlabeled training data in the video surveillance scenario. In this paper, we propose a new semi-supervised learning method called SIAVC for industrial accident video classification. Specifically, we design a video augmentation module called the Super Augmentation Block (SAB). SAB adds Gaussian noise and randomly masks video frames according to historical loss on the unlabeled data for model optimization. Then, we propose a Video Cross-set Augmentation Module (VCAM) to generate diverse pseudo-label samples from the high-confidence unlabeled samples, which alleviates the mismatch of sampling experience and provides high-quality training data. Additionally, we construct a new industrial accident surveillance video dataset with frame-level annotation, namely ECA9, to evaluate our proposed method. Compared with the state-of-the-art semi-supervised learning based methods, SIAVC demonstrates outstanding video classification performance, achieving 88.76% and 89.13% accuracy on ECA9 and Fire Detection datasets, respectively. The source code and the constructed dataset ECA9 will be released in https://github.com/AlchemyEmperor/SIAVC.
- 13 September 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 260
Leukocyte classification using relative-relationship-guided contrastive learning
- 13 September 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 260
Leukocyte classification using relative-relationship-guided contrastive learning
- 12 February 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 248
FireMatch: A semi-supervised video fire detection network based on consistency and distribution alignment
- 12 February 2024
- Published by Elsevier in Expert Systems with Applications
- Vol. 248
FireMatch: A semi-supervised video fire detection network based on consistency and distribution alignment