He-Feng Yin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
He-Feng Yin has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Xiao-Jun Wu
Jiangnan University
16 shared publications Xiao-Jun Wu
9 shared publications- Ziqi Li
School of Automation, Wuxi University, Wuxi, China
6 shared publications - Josef Kittler
University of Surrey
4 shared publications - Feng Jiang
School of Automation, Wuxi University, No. 333, Xishan Avenue Wuxi, Jiangsu 214105, P. R. China
3 shared publications Josef Kittler
3 shared publications
Latest Publications
- Open Access
- 5 November 2026
- Published by MDPI AG in Electronics
- Vol. 15 (21) , 5146
Dynamic Frequency Separated Fusion Network for RGB-T Semantic Segmentation
RGB-T semantic segmentation fuses visible (RGB) and thermal infrared (TIR) images to enhance model robustness under complex lighting conditions. However, most existing methods employ unified fusion strategies in the spatial domain, limiting their ability to fully exploit the complementary characteristics of the two modalities across different frequency components. To address this issue, we propose a novel fusion model that achieves finer cross-modal collaboration through frequency-aware feature disentanglement and adaptive fusion mechanisms. Specifically, the model first enhances the multi-granularity representation of dual-modality features using attention mechanisms. The features are then transformed into the frequency domain, where a dynamic band disentanglement mechanism adaptively separates high-frequency details from low-frequency semantics based on the input content. A heterogeneous fusion pathway is further designed to differentially aggregate high- and low-frequency components, emphasizing cross-modal detail alignment in the high-frequency path and semantic consistency in the low-frequency path. The proposed model is evaluated on three widely-used RGB-T segmentation benchmarks, namely MFNet, PST900, and FMB, and achieves state-of-the-art performance, demonstrating its effectiveness and superiority.
- 5 November 2026
- Published by MDPI AG in Electronics
- Vol. 15 (21) , 5146
Open AccessDynamic Frequency Separated Fusion Network for RGB-T Semantic Segmentation
RGB-T semantic segmentation fuses visible (RGB) and thermal infrared (TIR) images to enhance model robustness under complex lighting conditions. However, most existing methods employ unified fusion strategies in the spatial domain, limiting their ability to fully exploit the complementary characteristics of the two modalities across different frequency components. To address this issue, we propose a novel fusion model that achieves finer cross-modal collaboration through frequency-aware feature disentanglement and adaptive fusion mechanisms. Specifically, the model first enhances the multi-granularity representation of dual-modality features using attention mechanisms. The features are then transformed into the frequency domain, where a dynamic band disentanglement mechanism adaptively separates high-frequency details from low-frequency semantics based on the input content. A heterogeneous fusion pathway is further designed to differentially aggregate high- and low-frequency components, emphasizing cross-modal detail alignment in the high-frequency path and semantic consistency in the low-frequency path. The proposed model is evaluated on three widely-used RGB-T segmentation benchmarks, namely MFNet, PST900, and FMB, and achieves state-of-the-art performance, demonstrating its effectiveness and superiority.
- 9 July 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Wavelets, Multiresolution and Information Processing
- Vol. 24 (05)
Locality-sensitive competitive matrix regression for image classification
Over the past few years, nuclear norm-based matrix regression (NMR) has achieved superior performance in the field of image recognition. By employing the nuclear norm as the loss metric to preserve the two-dimensional structural information of error images, NMR exhibits strong robustness against structural noise. However, NMR fails to effectively exploit the category information of training data and ignores the local structural relationships among samples. To overcome the above limitations, we design a new classification method named locality-sensitive competitive matrix regression (LSCMR). First, a competitive representation term is introduced into the objective function of NMR, enforcing training samples from distinct categories to competitively characterize the test sample. To further boost the discriminative ability of representation coefficients, LSCMR incorporates the local structural information between samples, which encourages samples of the correct class to contribute more to the representation process. The alternating direction method of multipliers (ADMM) is utilized to solve the objective function of LSCMR. Experimental results on several public datasets verify the superiority of LSCMR. Specifically, on subset 4 and subset 5 of the Extended Yale B dataset, LSCMR achieves recognition accuracy of 97.37% and 62.18%, respectively, which are 9.21% and 20.02% higher than those of NMR. The source code of LSCMR is available at https://github.com/yinhefeng/LSCMR.
- 9 July 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Wavelets, Multiresolution and Information Processing
- Vol. 24 (05)
Locality-sensitive competitive matrix regression for image classification
Over the past few years, nuclear norm-based matrix regression (NMR) has achieved superior performance in the field of image recognition. By employing the nuclear norm as the loss metric to preserve the two-dimensional structural information of error images, NMR exhibits strong robustness against structural noise. However, NMR fails to effectively exploit the category information of training data and ignores the local structural relationships among samples. To overcome the above limitations, we design a new classification method named locality-sensitive competitive matrix regression (LSCMR). First, a competitive representation term is introduced into the objective function of NMR, enforcing training samples from distinct categories to competitively characterize the test sample. To further boost the discriminative ability of representation coefficients, LSCMR incorporates the local structural information between samples, which encourages samples of the correct class to contribute more to the representation process. The alternating direction method of multipliers (ADMM) is utilized to solve the objective function of LSCMR. Experimental results on several public datasets verify the superiority of LSCMR. Specifically, on subset 4 and subset 5 of the Extended Yale B dataset, LSCMR achieves recognition accuracy of 97.37% and 62.18%, respectively, which are 9.21% and 20.02% higher than those of NMR. The source code of LSCMR is available at https://github.com/yinhefeng/LSCMR.
- 23 April 2026
- Published by Elsevier in Pattern Recognition
- Vol. 179
Harmonising adaptive receptive field with multi-scale and multi-task perception for real-time face detection
- 23 April 2026
- Published by Elsevier in Pattern Recognition
- Vol. 179
Harmonising adaptive receptive field with multi-scale and multi-task perception for real-time face detection
- 25 March 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Wavelets, Multiresolution and Information Processing
- Vol. 24 (03)
A Weighted Competitive Nonnegative Representation Method for Image Classification
We present a weighted competitive nonnegative representation (WCNR) method. Specifically, WCNR introduces a competitive term, which leverages the competitive ability of training data from distinct classes to strengthen the connection between the representation and classification phases. In addition, WCNR incorporates a weight constraint. The weight constraint of each class imposed on the representation coefficients endows similar classes with more representation contributions, which boosts the discriminative power of the representation coefficients. To assess the classification performance of WCNR, extensive experiments are carried out on benchmarking datasets. Experimental results confirm that WCNR exceeds the state-of-the-art representation-based classification methods (RBCM) and also surpasses several deep learning approaches. The MATLAB code of WCNR is available at https://github.com/yinhefeng/WCNR .
- 25 March 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Wavelets, Multiresolution and Information Processing
- Vol. 24 (03)
A Weighted Competitive Nonnegative Representation Method for Image Classification
We present a weighted competitive nonnegative representation (WCNR) method. Specifically, WCNR introduces a competitive term, which leverages the competitive ability of training data from distinct classes to strengthen the connection between the representation and classification phases. In addition, WCNR incorporates a weight constraint. The weight constraint of each class imposed on the representation coefficients endows similar classes with more representation contributions, which boosts the discriminative power of the representation coefficients. To assess the classification performance of WCNR, extensive experiments are carried out on benchmarking datasets. Experimental results confirm that WCNR exceeds the state-of-the-art representation-based classification methods (RBCM) and also surpasses several deep learning approaches. The MATLAB code of WCNR is available at https://github.com/yinhefeng/WCNR .
- Open Access
- 25 February 2026
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 12 (3)
Affine non-negative collaborative representation based pattern classification
During the past decade, representation-based classification methods have received considerable attention in pattern recognition. In particular, the recently proposed non-negative representation based classification (NRC) method has been reported to achieve promising results in a wide range of classification tasks. However, NRC has two major drawbacks. First, there is no regularization term in the formulation of NRC, which may result in unstable solution and misclassification. Second, NRC ignores the fact that data usually lies in a union of multiple affine subspaces, rather than linear subspaces in practical applications. To address the above issues, this paper presents an affine non-negative collaborative representation (ANCR) model for pattern classification. To be more specific, ANCR imposes a regularization term on the coding vector. Moreover, since each point in an affine subspace can be expressed as an affine combination of other points in this affine subspace, ANCR introduces an affine constraint to better represent the data from affine subspaces. The experimental results on several benchmarking datasets demonstrate the merits of the proposed ANCR method. Concretely, on the Hopkins and Aircraft datasets, ANCR achieves accuracy of 97.8% and 87.7%, respectively, which represents an improvement of 2.2% and 0.4% over NRC. The source code of our ANCR is publicly available at https://github.com/yinhefeng/ANCR.
- 25 February 2026
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 12 (3)
Open AccessAffine non-negative collaborative representation based pattern classification
During the past decade, representation-based classification methods have received considerable attention in pattern recognition. In particular, the recently proposed non-negative representation based classification (NRC) method has been reported to achieve promising results in a wide range of classification tasks. However, NRC has two major drawbacks. First, there is no regularization term in the formulation of NRC, which may result in unstable solution and misclassification. Second, NRC ignores the fact that data usually lies in a union of multiple affine subspaces, rather than linear subspaces in practical applications. To address the above issues, this paper presents an affine non-negative collaborative representation (ANCR) model for pattern classification. To be more specific, ANCR imposes a regularization term on the coding vector. Moreover, since each point in an affine subspace can be expressed as an affine combination of other points in this affine subspace, ANCR introduces an affine constraint to better represent the data from affine subspaces. The experimental results on several benchmarking datasets demonstrate the merits of the proposed ANCR method. Concretely, on the Hopkins and Aircraft datasets, ANCR achieves accuracy of 97.8% and 87.7%, respectively, which represents an improvement of 2.2% and 0.4% over NRC. The source code of our ANCR is publicly available at https://github.com/yinhefeng/ANCR.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 141084-141094
HCCL: Hierarchical Consistency-Driven Contrastive Learning for Semi-Supervised Learning
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 141084-141094
Open AccessHCCL: Hierarchical Consistency-Driven Contrastive Learning for Semi-Supervised Learning
- Open Access
- 1 November 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (12) , 1-14
Pattern classification based on competitive and non-negative representation
- 1 November 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (12) , 1-14
Open AccessPattern classification based on competitive and non-negative representation
- Open Access
- 1 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 158237-158249
Fine-Grained Underwater Visual Object Tracking via Long-Term Template Refinement
With the growing demand for underwater recognition systems, underwater visual object tracking has recently received wide attention. The visible light scene analysis in a liquid environment poses specific challenges for underwater visual object tracking, such as impairment of sharpness, blue tinge, and illumination refraction. These problems lead to a performance degradation of the conventional visual object tracking algorithms. In this paper, we propose to adapt the methodology for underwater tracking by performing a long-term template refinement to achieve fine-grained spatial template matching and long-range temporal dependency for use with a Siamese tracking network. The template refinement is based on an attention mechanism, which is designed to highlight the saliency and discriminatory information of the tracked content. This mechanism obtains an accurate target mask for the template and alleviates the impact of the underwater noise in the template feature embedding. The long-term template generator updates the dynamic appearance of the target adaptively by clustering historical templates using a Gaussian Mixture Model. This approach enhances the target representation capacity and alleviates challenges such as visual obstructions, complex background variations and prevalent noise in underwater environment. The experimental results obtained on the challenging underwater tracking dataset, UOT32, demonstrate the advantage of the proposed long-term template refinement tracker. It outperforms state-of-the-art approaches such as SiamMask by 5.03% in Distance Precision and 2.31% in Overlap Success, which provides a promising visual perception technology for underwater artificial intelligence systems. On the standard tracking benchmarking dataset VOT2018, the proposed method is 5.56% higher than SiamRPN++ on the Expected Average Overlap and demonstrates comparable performance, verifying its generalisation capacity in other tracking scenarios and environments.
- 1 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 158237-158249
Open AccessFine-Grained Underwater Visual Object Tracking via Long-Term Template Refinement
With the growing demand for underwater recognition systems, underwater visual object tracking has recently received wide attention. The visible light scene analysis in a liquid environment poses specific challenges for underwater visual object tracking, such as impairment of sharpness, blue tinge, and illumination refraction. These problems lead to a performance degradation of the conventional visual object tracking algorithms. In this paper, we propose to adapt the methodology for underwater tracking by performing a long-term template refinement to achieve fine-grained spatial template matching and long-range temporal dependency for use with a Siamese tracking network. The template refinement is based on an attention mechanism, which is designed to highlight the saliency and discriminatory information of the tracked content. This mechanism obtains an accurate target mask for the template and alleviates the impact of the underwater noise in the template feature embedding. The long-term template generator updates the dynamic appearance of the target adaptively by clustering historical templates using a Gaussian Mixture Model. This approach enhances the target representation capacity and alleviates challenges such as visual obstructions, complex background variations and prevalent noise in underwater environment. The experimental results obtained on the challenging underwater tracking dataset, UOT32, demonstrate the advantage of the proposed long-term template refinement tracker. It outperforms state-of-the-art approaches such as SiamMask by 5.03% in Distance Precision and 2.31% in Overlap Success, which provides a promising visual perception technology for underwater artificial intelligence systems. On the standard tracking benchmarking dataset VOT2018, the proposed method is 5.56% higher than SiamRPN++ on the Expected Average Overlap and demonstrates comparable performance, verifying its generalisation capacity in other tracking scenarios and environments.
- 11 July 2025
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 28 (3) , 1-17
Affine locality-sensitive nonnegative representation based image classification
- 11 July 2025
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 28 (3) , 1-17
Affine locality-sensitive nonnegative representation based image classification
- 20 April 2025
- Published by Elsevier in Measurement
- Vol. 253
Research on gear defect detection using electromagnetic induction eddy current and infrared thermography
- 20 April 2025
- Published by Elsevier in Measurement
- Vol. 253
Research on gear defect detection using electromagnetic induction eddy current and infrared thermography