Zhiqun Pan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhiqun Pan has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Yongxiong Wang
School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, China
10 shared publications - Jiapeng Zhang
Fishery Machinery and Instrument Research Institute, Chinese Academy of Fishery Sciences, Shanghai 200092, China
5 shared publications - Xiaoming Wang
School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, China
4 shared publications - Guangpeng Wang
School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, China
2 shared publications Jiapeng Zhang
2 shared publications- Yiheng Tang
The School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai, China
2 shared publications
Latest Publications
- 13 April 2026
- Published by Elsevier in Neurocomputing
- Vol. 685
RRQ: Relative residual potential functions with knowledge-distilled value decomposition for MARL
- 13 April 2026
- Published by Elsevier in Neurocomputing
- Vol. 685
RRQ: Relative residual potential functions with knowledge-distilled value decomposition for MARL
- Open Access
- 17 June 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (8) , 1-23
DepthRL: a weakly supervised approach for monocular depth estimation using deep reinforcement learning
- 17 June 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (8) , 1-23
Open AccessDepthRL: a weakly supervised approach for monocular depth estimation using deep reinforcement learning
- 12 February 2025
- Published by Elsevier in Information Sciences
- Vol. 704
Distilled fine-grained domain adversarial network prompted by normalization and regularization
- 12 February 2025
- Published by Elsevier in Information Sciences
- Vol. 704
Distilled fine-grained domain adversarial network prompted by normalization and regularization
- 29 July 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (29) , 18499-18514
Unsupervised anomaly detection and localization via bidirectional knowledge distillation
- 29 July 2024
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (29) , 18499-18514
Unsupervised anomaly detection and localization via bidirectional knowledge distillation
- Open Access
- 27 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 49249-49260
A Double Self-Supervised Model for Pitting Detection on Ball Screws
Automatic detection of pitting on Ball Screw Drive (BSD) is essential to ensure normal production activities. However, the scarcity of defective samples and precisely labeled data poses a significant challenge. To address this, we propose an efficient double self-supervised model that operates at both the image and pixel levels, aiming to construct a high-performance model trained with defect-free data for detecting unknown defects in BSD images. By incorporating global and local information and extracting features at multiple hierarchical levels, the model’s generalization performance is enhanced. The image-level self-supervised representation is first learned by classifying normal images from the PasteNoise, a data augmentation approach by pasting noise patches at random locations in normal images. Meanwhile, the pixel-level self-supervised representation is learned by segmenting the noise patch to locate abnormal regions. Then, we introduce a novel feature masking strategy in a masking and prediction task for accurate defect localization. In addition, we use Histogram of Oriented Gradients (HOG) features with local contrast normalization as prediction targets to capture local shapes and appearances to improve the robustness of the model. The proposed method achieves competitive receiver operating characteristic curves of 97.42 (image-level) and 94.57 (pixel-level) on the BSD dataset. In experiments on the MVTec AD, the proposed model shows good performance, indicating the broad adaptability of our approach.
- 27 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 49249-49260
Open AccessA Double Self-Supervised Model for Pitting Detection on Ball Screws
Automatic detection of pitting on Ball Screw Drive (BSD) is essential to ensure normal production activities. However, the scarcity of defective samples and precisely labeled data poses a significant challenge. To address this, we propose an efficient double self-supervised model that operates at both the image and pixel levels, aiming to construct a high-performance model trained with defect-free data for detecting unknown defects in BSD images. By incorporating global and local information and extracting features at multiple hierarchical levels, the model’s generalization performance is enhanced. The image-level self-supervised representation is first learned by classifying normal images from the PasteNoise, a data augmentation approach by pasting noise patches at random locations in normal images. Meanwhile, the pixel-level self-supervised representation is learned by segmenting the noise patch to locate abnormal regions. Then, we introduce a novel feature masking strategy in a masking and prediction task for accurate defect localization. In addition, we use Histogram of Oriented Gradients (HOG) features with local contrast normalization as prediction targets to capture local shapes and appearances to improve the robustness of the model. The proposed method achieves competitive receiver operating characteristic curves of 97.42 (image-level) and 94.57 (pixel-level) on the BSD dataset. In experiments on the MVTec AD, the proposed model shows good performance, indicating the broad adaptability of our approach.
- 22 January 2024
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 132
Cross-domain self-supervised few-shot learning via multiple crops with teacher-student network
- 22 January 2024
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 132
Cross-domain self-supervised few-shot learning via multiple crops with teacher-student network
- Open Access
- 19 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 11836-11849
ViTFSL-Baseline: A Simple Baseline of Vision Transformer Network for Few-Shot Image Classification
Few-shot image classification, whose goal is to generalize to unseen tasks with scarce labeled data, has developed rapidly over the years. However, in traditional few-shot learning methods with CNNs, non-local features and long-rang dependencies of the image may be lost, and this leads to a poor generalization of the trained model. With the advantage of the self-attention mechanism of Transformer, researchers have tried to use vision transformer to improve few-shot learning recently. However, these methods are more complicated and take up a lot of computing resources, and there is no baseline to measure their effectiveness. We propose a new method called ViTFSL-baseline. We take advantage of vision transformer and train our model on all train set without episodic training. Meanwhile, we design a new nearest-neighbor classifier to used for few-shot image classification. Furthermore, in order to narrow the gap between difference of same class, we introduce centroid calibration in classifier after the feature extraction of backbone. We run the experiments on popular benchmarks to show that our method is a simple and effective for few-shot image classification. Our approach could be taken as the baseline upon vision transformer for few-shot learning.
- 19 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 11836-11849
Open AccessViTFSL-Baseline: A Simple Baseline of Vision Transformer Network for Few-Shot Image Classification
Few-shot image classification, whose goal is to generalize to unseen tasks with scarce labeled data, has developed rapidly over the years. However, in traditional few-shot learning methods with CNNs, non-local features and long-rang dependencies of the image may be lost, and this leads to a poor generalization of the trained model. With the advantage of the self-attention mechanism of Transformer, researchers have tried to use vision transformer to improve few-shot learning recently. However, these methods are more complicated and take up a lot of computing resources, and there is no baseline to measure their effectiveness. We propose a new method called ViTFSL-baseline. We take advantage of vision transformer and train our model on all train set without episodic training. Meanwhile, we design a new nearest-neighbor classifier to used for few-shot image classification. Furthermore, in order to narrow the gap between difference of same class, we introduce centroid calibration in classifier after the feature extraction of backbone. We run the experiments on popular benchmarks to show that our method is a simple and effective for few-shot image classification. Our approach could be taken as the baseline upon vision transformer for few-shot learning.
- Open Access
A Lightweight Domain Adversarial Neural Network Based on Knowledge Distillation for EEG-based Cross-subject Emotion Recognition
Individual differences of Electroencephalogram (EEG) could cause the domain shift which would significantly degrade the performance of cross-subject strategy. The domain adversarial neural networks (DANN), where the classification loss and domain loss jointly update the parameters of feature extractor, are adopted to deal with the domain shift. However, limited EEG data quantity and strong individual difference are challenges for the DANN with cumbersome feature extractor. In this work, we propose knowledge distillation (KD) based lightweight DANN to enhance cross-subject EEG-based emotion recognition. Specifically, the teacher model with strong context learning ability is utilized to learn complex temporal dynamics and spatial correlations of EEG, and robust lightweight student model is guided by the teacher model to learn more difficult domain-invariant features. In the feature-based KD framework, a transformer-based hierarchical temporalspatial learning model is served as the teacher model. The student model, which is composed of Bi-LSTM units, is a lightweight version of the teacher model. Hence, the student model could be supervised to mimic the robust feature representations of teacher model by leveraging complementary latent temporal features and spatial features. In the DANN-based cross-subject emotion recognition, we combine the obtained student model and a lightweight temporal-spatial feature interaction module as the feature extractor. And the feature aggregation is fed to the emotion classifier and domain classifier for domain-invariant feature learning. To verify the effectiveness of the proposed method, we conduct the subject-independent experiments on the public dataset DEAP with arousal and valence classification. The outstanding performance and t-SNE visualization of latent features verify the advantage and effectiveness of the proposed method.
Open AccessA Lightweight Domain Adversarial Neural Network Based on Knowledge Distillation for EEG-based Cross-subject Emotion Recognition
Individual differences of Electroencephalogram (EEG) could cause the domain shift which would significantly degrade the performance of cross-subject strategy. The domain adversarial neural networks (DANN), where the classification loss and domain loss jointly update the parameters of feature extractor, are adopted to deal with the domain shift. However, limited EEG data quantity and strong individual difference are challenges for the DANN with cumbersome feature extractor. In this work, we propose knowledge distillation (KD) based lightweight DANN to enhance cross-subject EEG-based emotion recognition. Specifically, the teacher model with strong context learning ability is utilized to learn complex temporal dynamics and spatial correlations of EEG, and robust lightweight student model is guided by the teacher model to learn more difficult domain-invariant features. In the feature-based KD framework, a transformer-based hierarchical temporalspatial learning model is served as the teacher model. The student model, which is composed of Bi-LSTM units, is a lightweight version of the teacher model. Hence, the student model could be supervised to mimic the robust feature representations of teacher model by leveraging complementary latent temporal features and spatial features. In the DANN-based cross-subject emotion recognition, we combine the obtained student model and a lightweight temporal-spatial feature interaction module as the feature extractor. And the feature aggregation is fed to the emotion classifier and domain classifier for domain-invariant feature learning. To verify the effectiveness of the proposed method, we conduct the subject-independent experiments on the public dataset DEAP with arousal and valence classification. The outstanding performance and t-SNE visualization of latent features verify the advantage and effectiveness of the proposed method.
- 27 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 19 (12) , 11675-11686
Dual-Branch TransV-Net for 3-D Echocardiography Segmentation
Segmentation of left and right ventricles from 3D echocardiographic images is premised and key for quantitative analysis of cardiac function, which is important for pediatric cardiac diagnosis. Compared with 2D echocardiography, 3D echocardiography can fully represent ventricular structure without geometric inference. However, it usually takes experts several hours to obtain a 3D segmentation mask of the left and right ventricles. Therefore, a fast and automatic segmentation method is highly desired. Unfortunately, 3D echocardiography suffers from low contrast, unclear left and right ventricles borders, and blind zone. Existing segmentation methods usually have poor performance at ventricular boundaries. To deal with these problems, we propose a novel dual-branch TransV-Net (DBTV). DBTV comprises two parallel, interleaved, and relatively independent V-shaped encoder-decoder branches. The main branch acts on the original data to extract image features, and the auxiliary branch acts on edge maps to extract the additional edge features. To suppress noise and enhance the edge information, extra concatenations are added to bridge the features from the main and auxiliary branches. To reduce object missing caused by blind zone, a 3D transformer-based module is proposed in the bottom layer of the dual-branch structure to extract the global contexts. We do experiments on a self-collected dataset with 120 3D echocardiographic images from 60 cardiac sequences, and the dice score of 0.913 and 0.880 are obtained in the left and right ventricle segments, respectively. Each inference takes about two-thirds of a second for a single 3D frame.
- 27 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 19 (12) , 11675-11686
Dual-Branch TransV-Net for 3-D Echocardiography Segmentation
Segmentation of left and right ventricles from 3D echocardiographic images is premised and key for quantitative analysis of cardiac function, which is important for pediatric cardiac diagnosis. Compared with 2D echocardiography, 3D echocardiography can fully represent ventricular structure without geometric inference. However, it usually takes experts several hours to obtain a 3D segmentation mask of the left and right ventricles. Therefore, a fast and automatic segmentation method is highly desired. Unfortunately, 3D echocardiography suffers from low contrast, unclear left and right ventricles borders, and blind zone. Existing segmentation methods usually have poor performance at ventricular boundaries. To deal with these problems, we propose a novel dual-branch TransV-Net (DBTV). DBTV comprises two parallel, interleaved, and relatively independent V-shaped encoder-decoder branches. The main branch acts on the original data to extract image features, and the auxiliary branch acts on edge maps to extract the additional edge features. To suppress noise and enhance the edge information, extra concatenations are added to bridge the features from the main and auxiliary branches. To reduce object missing caused by blind zone, a 3D transformer-based module is proposed in the bottom layer of the dual-branch structure to extract the global contexts. We do experiments on a self-collected dataset with 120 3D echocardiographic images from 60 cardiac sequences, and the dice score of 0.913 and 0.880 are obtained in the left and right ventricle segments, respectively. Each inference takes about two-thirds of a second for a single 3D frame.