Wenting Shen
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
Research Subjects
Wenting Shen has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Jiazheng Wang
Hunan University
6 shared publications - Min Liu
Hunan University
6 shared publications - Yaonan Wang
Hunan University
6 shared publications - Erik Meijering
School of Computer Science and Engineering, University of New South Wales, Sydney, Australia
4 shared publications - Renjie Ding
College of Electrical and Information Engineering, the National Engineering Research Center of Robot Visual Perception and Control Technology, and the International Scientific and Technological Innovation Cooperation Base for Biomedical Image Processing, Hunan University, Changsha, Hunan, China
4 shared publications - Zhe Zhang
Guangxi University
3 shared publications
Latest Publications
- 9 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
Exploring Volume Representation Similarity in Long-Tail Biased Stereo Matching
Stereo matching is fundamental to many vision applications. However, prior work has largely centered on network architectures while overlooking the inherent long-tail bias and its influence on the pipeline. Tail targets carry critical information, and analyzing the training pipeline with bias is essential for understanding network behavior and improving performance. In this paper, we introduce rebalancing techniques and a volume representation with a similarity metric, tailored for stereo-matching networks, to study the biased influence of pretraining on subsequent finetuning. Building on this foundation, we first re-examine existing networks from four perspectives: model comparison, dual-stage training, data augmentation, and regression analysis. Our results show that the some gains of previous methods may stem from long-tail bias, where overall improvements coincide with degraded performance on the tail. We further explore how similarity changes within the specific biased pipeline and demonstrate that it is a meaningful indicator of how well pretraining benefits finetuning: higher similarity during pretraining consistently yields better finetuning outcomes. To validate this insight, we propose an improved loss that explicitly increases representation similarity in pretraining, leading to stronger finetuning performance. Experiments support our hypotheses, highlighting the value of representation similarity and offering new perspectives for advancing stereo matching research. Code is available at https://github.com/EliottDJay/ImbStereo.
- 9 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
Exploring Volume Representation Similarity in Long-Tail Biased Stereo Matching
Stereo matching is fundamental to many vision applications. However, prior work has largely centered on network architectures while overlooking the inherent long-tail bias and its influence on the pipeline. Tail targets carry critical information, and analyzing the training pipeline with bias is essential for understanding network behavior and improving performance. In this paper, we introduce rebalancing techniques and a volume representation with a similarity metric, tailored for stereo-matching networks, to study the biased influence of pretraining on subsequent finetuning. Building on this foundation, we first re-examine existing networks from four perspectives: model comparison, dual-stage training, data augmentation, and regression analysis. Our results show that the some gains of previous methods may stem from long-tail bias, where overall improvements coincide with degraded performance on the tail. We further explore how similarity changes within the specific biased pipeline and demonstrate that it is a meaningful indicator of how well pretraining benefits finetuning: higher similarity during pretraining consistently yields better finetuning outcomes. To validate this insight, we propose an improved loss that explicitly increases representation similarity in pretraining, leading to stronger finetuning performance. Experiments support our hypotheses, highlighting the value of representation similarity and offering new perspectives for advancing stereo matching research. Code is available at https://github.com/EliottDJay/ImbStereo.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 10 (3) , 2498-2515
Weak-Strong Consistency Learning via Collaborative Space Alignment for Semi-Supervised Medical Image Segmentation
In semi-supervised learning, cross pseudo supervision (CPS) is considered a promising learning approach.However, we observe that CPS easily converges to consensus early in training and degenerates into noisy self-training. To tackle this challenge, we propose a novel semi-supervised segmentation framework that utilizes collaborative spatial complementary alignment to facilitate weak-to-strong consistency learning (CSWS). CSWS comprises two main components: Weak-to-Strong Consistency Learning Strategy (WSCL) and Collaborative Space Complementary Alignment Strategy (CSCA). WSCL is implemented by combining weak augmentation and strong augmentation in a differently initialized network, thus effectively combining the benefits of consistency and self-training. To suppress the impact of low-quality pseudo-labels on consistency learning, CSCA leverages the semantic information in logit space and representation space to offer complementary features for consistency learning in a collaborative manner. In addition, we enrich the model inference view differences by introducing a prototype feature discrepancy loss, which is expected to facilitate consistent learning to obtain more precise predictions. We have conducted extensive experiments on five public datasets, including MRI, CT, and endoscopic imaging modalities. The experimental results show that CSWS has better generalization and high robustness than other state-of-the-art methods.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 10 (3) , 2498-2515
Weak-Strong Consistency Learning via Collaborative Space Alignment for Semi-Supervised Medical Image Segmentation
In semi-supervised learning, cross pseudo supervision (CPS) is considered a promising learning approach.However, we observe that CPS easily converges to consensus early in training and degenerates into noisy self-training. To tackle this challenge, we propose a novel semi-supervised segmentation framework that utilizes collaborative spatial complementary alignment to facilitate weak-to-strong consistency learning (CSWS). CSWS comprises two main components: Weak-to-Strong Consistency Learning Strategy (WSCL) and Collaborative Space Complementary Alignment Strategy (CSCA). WSCL is implemented by combining weak augmentation and strong augmentation in a differently initialized network, thus effectively combining the benefits of consistency and self-training. To suppress the impact of low-quality pseudo-labels on consistency learning, CSCA leverages the semantic information in logit space and representation space to offer complementary features for consistency learning in a collaborative manner. In addition, we enrich the model inference view differences by introducing a prototype feature discrepancy loss, which is expected to facilitate consistent learning to obtain more precise predictions. We have conducted extensive experiments on five public datasets, including MRI, CT, and endoscopic imaging modalities. The experimental results show that CSWS has better generalization and high robustness than other state-of-the-art methods.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 6184-6196
UniSurg: A Unified Multitask Framework for Robotic Surgical Scene Understanding
Surgical scene understanding is a vital intelligent technique in robot-assisted surgery, including surgical instrument detection, segmentation, and instrument–tissue interaction detection. Existing methods typically address these tasks in isolation, neglecting the intrinsic correlations among them. In this work, we innovatively propose a unified multitask framework named UniSurg, being the first to jointly address these three critical aspects of surgical scene understanding, thereby providing the robot with multidimensional perceptual capabilities. By exploring the inter-task correlations and reusing shared features, UniSurg has been demonstrated to significantly enhance the scene analysis performance. To address pose variability of the instruments under the constrained field of view in laparoscopic surgery, we design an Attention Enhanced Conditional Convolution (AEC-Conv) that dynamically adjusts kernels based on pose-specific features for improved adaptability. To further enhance interaction detection, we propose the Temporal Difference Enhancement module (TDE), which captures motion cues by amplifying inter-frame differences, and the Pyramid Global Feature Enhancement module (PGFE), which leverages graph-based hierarchical context to model global relational dependencies. Experiments on the Endovis2018 dataset and a clinical multitask dataset MILVis demonstrate the superior multitask performance of UniSurg.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (5) , 6184-6196
UniSurg: A Unified Multitask Framework for Robotic Surgical Scene Understanding
Surgical scene understanding is a vital intelligent technique in robot-assisted surgery, including surgical instrument detection, segmentation, and instrument–tissue interaction detection. Existing methods typically address these tasks in isolation, neglecting the intrinsic correlations among them. In this work, we innovatively propose a unified multitask framework named UniSurg, being the first to jointly address these three critical aspects of surgical scene understanding, thereby providing the robot with multidimensional perceptual capabilities. By exploring the inter-task correlations and reusing shared features, UniSurg has been demonstrated to significantly enhance the scene analysis performance. To address pose variability of the instruments under the constrained field of view in laparoscopic surgery, we design an Attention Enhanced Conditional Convolution (AEC-Conv) that dynamically adjusts kernels based on pose-specific features for improved adaptability. To further enhance interaction detection, we propose the Temporal Difference Enhancement module (TDE), which captures motion cues by amplifying inter-frame differences, and the Pyramid Global Feature Enhancement module (PGFE), which leverages graph-based hierarchical context to model global relational dependencies. Experiments on the Endovis2018 dataset and a clinical multitask dataset MILVis demonstrate the superior multitask performance of UniSurg.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (1) , 379-390
EPDiff: Erasure Perception Diffusion Model for Unsupervised Anomaly Detection in Preoperative Multimodal Images
Unsupervised anomaly detection (UAD) methods typically detect anomalies by learning and reconstructing the normative distribution. However, since anomalies constantly invade and affect their surroundings, sub-healthy areas in the junction present structural deformations that could be easily misidentified as anomalies, posing difficulties for UAD methods that solely learn the normative distribution. The use of multimodal images can facilitate to address the above challenges, as they can provide complementary information of anomalies. Therefore, this paper propose a novel method for UAD in preoperative multimodal images, called Erasure Perception Diffusion model (EPDiff). First, the Local Erasure Progressive Training (LEPT) framework is designed to better rebuild sub-healthy structures around anomalies through the diffusion model with a two-phase process. Initially, healthy images are used to capture deviation features labeled as potential anomalies. Then, these anomalies are locally erased in multimodal images to progressively learn sub-healthy structures, obtaining a more detailed reconstruction around anomalies. Second, the Global Structural Perception (GSP) module is developed in the diffusion model to realize global structural representation and correlation within images and between modalities through interactions of high-level semantic information. In addition, a training-free module, named Multimodal Attention Fusion (MAF) module, is presented for weighted fusion of anomaly maps between different modalities and obtaining binary anomaly outputs. Experimental results show that EPDiff improves the AUPRC and mDice scores by 2% and 3.9% on BraTS2021, and by 5.2% and 4.5% on Shifts over the state-of-the-art methods, which proves the applicability of EPDiff in diverse anomaly diagnosis. The code is available at https://github.com/wjiazheng/EPDiff.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (1) , 379-390
EPDiff: Erasure Perception Diffusion Model for Unsupervised Anomaly Detection in Preoperative Multimodal Images
Unsupervised anomaly detection (UAD) methods typically detect anomalies by learning and reconstructing the normative distribution. However, since anomalies constantly invade and affect their surroundings, sub-healthy areas in the junction present structural deformations that could be easily misidentified as anomalies, posing difficulties for UAD methods that solely learn the normative distribution. The use of multimodal images can facilitate to address the above challenges, as they can provide complementary information of anomalies. Therefore, this paper propose a novel method for UAD in preoperative multimodal images, called Erasure Perception Diffusion model (EPDiff). First, the Local Erasure Progressive Training (LEPT) framework is designed to better rebuild sub-healthy structures around anomalies through the diffusion model with a two-phase process. Initially, healthy images are used to capture deviation features labeled as potential anomalies. Then, these anomalies are locally erased in multimodal images to progressively learn sub-healthy structures, obtaining a more detailed reconstruction around anomalies. Second, the Global Structural Perception (GSP) module is developed in the diffusion model to realize global structural representation and correlation within images and between modalities through interactions of high-level semantic information. In addition, a training-free module, named Multimodal Attention Fusion (MAF) module, is presented for weighted fusion of anomaly maps between different modalities and obtaining binary anomaly outputs. Experimental results show that EPDiff improves the AUPRC and mDice scores by 2% and 3.9% on BraTS2021, and by 5.2% and 4.5% on Shifts over the state-of-the-art methods, which proves the applicability of EPDiff in diverse anomaly diagnosis. The code is available at https://github.com/wjiazheng/EPDiff.
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Multiview Integration Network for Multitask Robotic Surgical Scene Analysis
Surgical scene analysis holds a pivotal role in robot-assisted surgery. However, existing methods often suffer from single or little views, leading to erroneous scene analysis conclusions. To address these issues, a novel Multiview Integration Network (MVINet) is proposed, which comprehensively analyzes the surgical scene by integrating effective information from multiple views, including global, local, dynamic, and static views. As a multitask scene analysis network, MVINet could simultaneously perform semantic segmentation of surgical instruments and detection of instrument-tissue interaction. By designing a unique Global-Local Spatial Feature Memory module (G-LSFM), the joint information from globally and locally analyzed views following graph analysis simultaneously enhances the accuracy of multitask scene analysis. The Dynamic-Static Visual Feature Memory module (D-SVFM) introduces an innovative approach by simultaneously incorporating temporal feature from consecutive frames and static feature from single frame into the node features of the interaction reasoning network. The multiscale perspective of both features further enhances the ability of the module to analyze the complex scenes. Experimental results on a public and a private dataset demonstrate that our method achieves superior performance compared to other state-of-the-art methods for two crucial tasks in surgical scene analysis. In the instrument segmentation task, MVINet outperforms the second-best method by 1.73% and 0.55% in terms of mIoU scores. For the interaction detection task, MVINet surpasses the second-best method by 10.14% and 5.47% in mAP scores.
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Multiview Integration Network for Multitask Robotic Surgical Scene Analysis
Surgical scene analysis holds a pivotal role in robot-assisted surgery. However, existing methods often suffer from single or little views, leading to erroneous scene analysis conclusions. To address these issues, a novel Multiview Integration Network (MVINet) is proposed, which comprehensively analyzes the surgical scene by integrating effective information from multiple views, including global, local, dynamic, and static views. As a multitask scene analysis network, MVINet could simultaneously perform semantic segmentation of surgical instruments and detection of instrument-tissue interaction. By designing a unique Global-Local Spatial Feature Memory module (G-LSFM), the joint information from globally and locally analyzed views following graph analysis simultaneously enhances the accuracy of multitask scene analysis. The Dynamic-Static Visual Feature Memory module (D-SVFM) introduces an innovative approach by simultaneously incorporating temporal feature from consecutive frames and static feature from single frame into the node features of the interaction reasoning network. The multiscale perspective of both features further enhances the ability of the module to analyze the complex scenes. Experimental results on a public and a private dataset demonstrate that our method achieves superior performance compared to other state-of-the-art methods for two crucial tasks in surgical scene analysis. In the instrument segmentation task, MVINet outperforms the second-best method by 1.73% and 0.55% in terms of mIoU scores. For the interaction detection task, MVINet surpasses the second-best method by 10.14% and 5.47% in mAP scores.
- 21 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 42 (11) , 3408-3419
Branch Aggregation Attention Network for Robotic Surgical Instrument Segmentation
Surgical instrument segmentation is of great significance to robot-assisted surgery, but the noise caused by reflection, water mist, and motion blur during the surgery as well as the different forms of surgical instruments would greatly increase the difficulty of precise segmentation. A novel method called Branch Aggregation Attention network (BAANet) is proposed to address these challenges, which adopts a lightweight encoder and two designed modules, named Branch Balance Aggregation module (BBA) and Block Attention Fusion module (BAF), for efficient feature localization and denoising. By introducing the unique BBA module, features from multiple branches are balanced and optimized through a combination of addition and multiplication to complement strengths and effectively suppress noise. Furthermore, to fully integrate the contextual information and capture the region of interest, the BAF module is proposed in the decoder, which receives adjacent feature maps from the BBA module and localizes the surgical instruments from both global and local perspectives by utilizing a dual branch attention mechanism. According to the experimental results, the proposed method has the advantage of being lightweight while outperforming the second-best method by 4.03%, 1.53%, and 1.34% in mIoU scores on three challenging surgical instrument datasets, respectively, compared to the existing state-of-the-art methods. Code is available at https://github.com/SWT-1014/BAANet.
- 21 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 42 (11) , 3408-3419
Branch Aggregation Attention Network for Robotic Surgical Instrument Segmentation
Surgical instrument segmentation is of great significance to robot-assisted surgery, but the noise caused by reflection, water mist, and motion blur during the surgery as well as the different forms of surgical instruments would greatly increase the difficulty of precise segmentation. A novel method called Branch Aggregation Attention network (BAANet) is proposed to address these challenges, which adopts a lightweight encoder and two designed modules, named Branch Balance Aggregation module (BBA) and Block Attention Fusion module (BAF), for efficient feature localization and denoising. By introducing the unique BBA module, features from multiple branches are balanced and optimized through a combination of addition and multiplication to complement strengths and effectively suppress noise. Furthermore, to fully integrate the contextual information and capture the region of interest, the BAF module is proposed in the decoder, which receives adjacent feature maps from the BBA module and localizes the surgical instruments from both global and local perspectives by utilizing a dual branch attention mechanism. According to the experimental results, the proposed method has the advantage of being lightweight while outperforming the second-best method by 4.03%, 1.53%, and 1.34% in mIoU scores on three challenging surgical instrument datasets, respectively, compared to the existing state-of-the-art methods. Code is available at https://github.com/SWT-1014/BAANet.
- Open Access
- 11 January 2022
- Published by Springer Nature in Communications Biology
- Vol. 5 (1) , 1-8
Independent control of amplitude and period in a synthetic oscillator circuit with modified repressilator
Synthetic Biology aims to create predictable biological circuits and fully operational biological systems. Although there are methods to create more stable oscillators, such as repressilators, independently controlling the oscillation of reporter genes in terms of their amplitude and period is only on theoretical level. Here, we introduce a new oscillator circuit that can be independently controlled by two inducers in Escherichia coli. Some control components, including σECF11 and NahR, were added to the circuit. By systematically tuning the concentration of the inducers, salicylate and IPTG, the amplitude and period can be modulated independently. Furthermore, we constructed a quantitative model to forecast the regulation results. Under the guidance of the model, the expected oscillation can be regulated by choosing the proper concentration combinations of inducers. In summary, our work achieved independent control of the oscillator circuit, which allows the oscillator to be modularized and used in more complex circuit designs.
- 11 January 2022
- Published by Springer Nature in Communications Biology
- Vol. 5 (1) , 1-8
Open AccessIndependent control of amplitude and period in a synthetic oscillator circuit with modified repressilator
Synthetic Biology aims to create predictable biological circuits and fully operational biological systems. Although there are methods to create more stable oscillators, such as repressilators, independently controlling the oscillation of reporter genes in terms of their amplitude and period is only on theoretical level. Here, we introduce a new oscillator circuit that can be independently controlled by two inducers in Escherichia coli. Some control components, including σECF11 and NahR, were added to the circuit. By systematically tuning the concentration of the inducers, salicylate and IPTG, the amplitude and period can be modulated independently. Furthermore, we constructed a quantitative model to forecast the regulation results. Under the guidance of the model, the expected oscillation can be regulated by choosing the proper concentration combinations of inducers. In summary, our work achieved independent control of the oscillator circuit, which allows the oscillator to be modularized and used in more complex circuit designs.