Linfei Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Linfei Wang has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Dapeng Tao
Yunnan University
4 shared publications - Yibing Zhan
Wuhan University
4 shared publications - Baosheng Yu
Nanyang Technological University
2 shared publications - Xu Lin
Yunnan Union Vision Innovation Technology Company Ltd., Kunming, Yunnan, China
2 shared publications - Yonghang Tai
Yunnan Normal University
2 shared publications - Changshun Song
Yunnan Normal University, Kunming, Yunnan, China
1 shared publication
Latest Publications
- 19 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-14
An Ultrasound-Guided Tumor Puncture Framework With Vascular Avoidance for Human–Machine Collaborative Biopsy
Ultrasound (US) imaging has gained widespread application in guiding biopsy procedures. However, conventional manual US-guided biopsy remains limited by occupational exposure risks, limited accuracy in vessel and lesion identification, and strong dependence on operator proficiency. To address these challenges, we propose a tumor biopsy framework centered on a tele-assisted robotic system. The proposed framework, based on visual servoing control with a Kalman filter–assisted PD strategy and a leader–follower human–machine collaboration scheme, achieves high-precision puncture around the incision site, thereby enabling real-time tumor localization and vascular avoidance. At the end effector, a two-degree-of-freedom (2-DOF) remote center-of-motion (RCM) mechanism enables concurrent operation of the US probe and biopsy needle within confined spaces. In addition, a geometric quantification strategy based on spatial calibration is introduced to support millimeter-level error assessment. Performance was evaluated in silicone phantoms across three puncture modes. Compared with conventional manual puncture, the robotic system achieved a mean positioning error of 2.00±0.54 mm (a 47.8% improvement) and an angular error of 4.81°±1.10° (a 36.3% improvement), meeting clinical requirements for biopsy guidance. These results indicate that the proposed framework can substantially enhance procedural accuracy while reducing dependence on operator skill and clinical experience, thereby holding promise for safer and more reliable US-guided biopsy. The Project link: https://github.com/LinfieWang/HMC-Biopsy.
- 19 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-14
An Ultrasound-Guided Tumor Puncture Framework With Vascular Avoidance for Human–Machine Collaborative Biopsy
Ultrasound (US) imaging has gained widespread application in guiding biopsy procedures. However, conventional manual US-guided biopsy remains limited by occupational exposure risks, limited accuracy in vessel and lesion identification, and strong dependence on operator proficiency. To address these challenges, we propose a tumor biopsy framework centered on a tele-assisted robotic system. The proposed framework, based on visual servoing control with a Kalman filter–assisted PD strategy and a leader–follower human–machine collaboration scheme, achieves high-precision puncture around the incision site, thereby enabling real-time tumor localization and vascular avoidance. At the end effector, a two-degree-of-freedom (2-DOF) remote center-of-motion (RCM) mechanism enables concurrent operation of the US probe and biopsy needle within confined spaces. In addition, a geometric quantification strategy based on spatial calibration is introduced to support millimeter-level error assessment. Performance was evaluated in silicone phantoms across three puncture modes. Compared with conventional manual puncture, the robotic system achieved a mean positioning error of 2.00±0.54 mm (a 47.8% improvement) and an angular error of 4.81°±1.10° (a 36.3% improvement), meeting clinical requirements for biopsy guidance. These results indicate that the proposed framework can substantially enhance procedural accuracy while reducing dependence on operator skill and clinical experience, thereby holding promise for safer and more reliable US-guided biopsy. The Project link: https://github.com/LinfieWang/HMC-Biopsy.
- Open Access
- 13 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 697-703
Real-Time Prony Series Haptic Rendering for Viscoelastic Lung Modeling in Virtual Surgery
- 13 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 697-703
Open AccessReal-Time Prony Series Haptic Rendering for Viscoelastic Lung Modeling in Virtual Surgery
- 4 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 1652-1664
Cps-STS: Bridging the Gap Between Content and Position for Coarse-Point-Supervised Scene Text Spotter
Recently, weakly supervised methods for scene text spotter are increasingly popular with researchers due to their potential to significantly reduce dataset annotation efforts. The latest progress in this field is text spotter based on single or multi-point annotations. However, this method struggles with the sensitivity of text recognition to the precise annotation location and fails to capture the relative positions and shapes of characters, leading to impaired recognition of texts with extensive rotations and flips. To address these challenges, this paper develops a novel method named Coarse-point-supervised Scene Text Spotter (Cps-STS). Cps-STS first utilizes a few approximate points as text location labels and introduces a learnable position modulation mechanism, easing the accuracy requirements for annotations and enhancing model robustness. Additionally, we incorporate a Spatial Compatibility Attention (SCA) module for text decoding to effectively utilize spatial data such as position and shape. This module fuses compound queries and global feature maps, serving as a bias in the SCA module to express text spatial morphology. In order to accurately locate and decode text content, we introduce features containing spatial morphology information and text content into the input features of the text decoder. By introducing features with spatial morphology information as bias terms into the text decoder, ablation experiments demonstrate that this operation enables the model to effectively identify and utilize the relationship between text content and position to enhance the recognition performance of our model. One significant advantage of Cps-STS is its ability to achieve full supervision-level performance with just a few imprecise coarse points at a low cost. Extensive experiments validate the effectiveness and superiority of Cps-STS over existing approaches.
- 4 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 1652-1664
Cps-STS: Bridging the Gap Between Content and Position for Coarse-Point-Supervised Scene Text Spotter
Recently, weakly supervised methods for scene text spotter are increasingly popular with researchers due to their potential to significantly reduce dataset annotation efforts. The latest progress in this field is text spotter based on single or multi-point annotations. However, this method struggles with the sensitivity of text recognition to the precise annotation location and fails to capture the relative positions and shapes of characters, leading to impaired recognition of texts with extensive rotations and flips. To address these challenges, this paper develops a novel method named Coarse-point-supervised Scene Text Spotter (Cps-STS). Cps-STS first utilizes a few approximate points as text location labels and introduces a learnable position modulation mechanism, easing the accuracy requirements for annotations and enhancing model robustness. Additionally, we incorporate a Spatial Compatibility Attention (SCA) module for text decoding to effectively utilize spatial data such as position and shape. This module fuses compound queries and global feature maps, serving as a bias in the SCA module to express text spatial morphology. In order to accurately locate and decode text content, we introduce features containing spatial morphology information and text content into the input features of the text decoder. By introducing features with spatial morphology information as bias terms into the text decoder, ablation experiments demonstrate that this operation enables the model to effectively identify and utilize the relationship between text content and position to enhance the recognition performance of our model. One significant advantage of Cps-STS is its ability to achieve full supervision-level performance with just a few imprecise coarse points at a low cost. Extensive experiments validate the effectiveness and superiority of Cps-STS over existing approaches.
- 17 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 34 (11) , 11086-11100
DeIoU: Toward Distinguishable Box Prediction in Densely Packed Object Detection
The Intersection over Union (IoU) has been widely employed in various stages of object detection owing to its ability to quantify the similarity between boxes objectively. However, in densely packed scenes full of crowded and small-sized objects, adjacent positive boxes often exhibit high levels of overlap. This overlap interference compromises the consistency between quality evaluation and confidence, leading to ambiguous box prediction within the previous IoU-based models. To address this issue, we design a novel learning paradigm tailored for Dense scenes based on IoU, called DeIoU. This approach effectively suppresses unnecessary overlap between predicted boxes and thereby enhances representation learning for non-salient objects. Specifically, it consists of a dense box regression loss LDeIoU and a one-to-many (O2M) label matching strategy guided by DeIoU. These components focus on calibrating the position and shape prediction quality during the model training, learning distinguishable object features by penalizing overlap interference between neighboring boxes. Extensive experiments on four object detection datasets including SKU-110K, CrowdHuman, MS COCO 2017, and DIOR, demonstrate that our DeIoU-based learning strategy outperforms other state-of-the-art methods. Notably, the proposed method delivers a substantial improvement (average 1.3 AP and 1.8 M R -2 ) across popular detectors on SKU-110K and CrowdHuman while exhibiting distinct competitiveness on small objects within natural scenes.
- 17 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 34 (11) , 11086-11100
DeIoU: Toward Distinguishable Box Prediction in Densely Packed Object Detection
The Intersection over Union (IoU) has been widely employed in various stages of object detection owing to its ability to quantify the similarity between boxes objectively. However, in densely packed scenes full of crowded and small-sized objects, adjacent positive boxes often exhibit high levels of overlap. This overlap interference compromises the consistency between quality evaluation and confidence, leading to ambiguous box prediction within the previous IoU-based models. To address this issue, we design a novel learning paradigm tailored for Dense scenes based on IoU, called DeIoU. This approach effectively suppresses unnecessary overlap between predicted boxes and thereby enhances representation learning for non-salient objects. Specifically, it consists of a dense box regression loss LDeIoU and a one-to-many (O2M) label matching strategy guided by DeIoU. These components focus on calibrating the position and shape prediction quality during the model training, learning distinguishable object features by penalizing overlap interference between neighboring boxes. Extensive experiments on four object detection datasets including SKU-110K, CrowdHuman, MS COCO 2017, and DIOR, demonstrate that our DeIoU-based learning strategy outperforms other state-of-the-art methods. Notably, the proposed method delivers a substantial improvement (average 1.3 AP and 1.8 M R -2 ) across popular detectors on SKU-110K and CrowdHuman while exhibiting distinct competitiveness on small objects within natural scenes.
- 20 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 9 (1) , 509-521
Explicit and Implicit Box Equivariance Learning for Weakly-Supervised Rotated Object Detection
Rotated object detection has emerged as a preferred alternative to the conventional Horizontal Box (HBox) object detection in various tasks. However, the majority of existing detection datasets are still annotated using HBox. Re-annotating these datasets for rotated object detection is notably inefficient and labor-intensive. Recently, weakly-supervised detectors based directly on HBox have provided a solution to bridge the gap between the rotated bounding box (RBox) prediction and the weak annotations. Nonetheless, they tend to overlook the potential impact of factors such as bounding box position, aspect ratio, and scale on angle prediction. Additionally, inconsistencies arise when evaluating rotation variations, hindering the inference performance. In light of these limitations, we propose an explicit and implicit equivariance detector (EIE-Det) to predict fine-grained RBox based on HBox annotations. Along with a shared backbone and neck structure, our proposed EIE-Det comprises a weakly supervised (WS) branch for predicting coarse RBox from HBox labels, an explicit equivariance (EE) branch for learning rotation consistency, and an implicit equivariance (IE) branch for learning position, aspect ratio, and scale consistency. Furthermore, we introduce the Tanimoto Coefficient for vector-based bounding box regression, enabling a comprehensive assessment of model prediction bias, thereby eliminating interference from variations in box aspect ratio and scale. Experiments demonstrate that our method not only significantly surpasses state-of-the-art models supervised with HBox but also achieves accuracy comparable to fully-supervised detectors (75.74$\%$ and 58.50$\%$ $AP_{50}$ on DOTA and DIOR-R, respectively).
- 20 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 9 (1) , 509-521
Explicit and Implicit Box Equivariance Learning for Weakly-Supervised Rotated Object Detection
Rotated object detection has emerged as a preferred alternative to the conventional Horizontal Box (HBox) object detection in various tasks. However, the majority of existing detection datasets are still annotated using HBox. Re-annotating these datasets for rotated object detection is notably inefficient and labor-intensive. Recently, weakly-supervised detectors based directly on HBox have provided a solution to bridge the gap between the rotated bounding box (RBox) prediction and the weak annotations. Nonetheless, they tend to overlook the potential impact of factors such as bounding box position, aspect ratio, and scale on angle prediction. Additionally, inconsistencies arise when evaluating rotation variations, hindering the inference performance. In light of these limitations, we propose an explicit and implicit equivariance detector (EIE-Det) to predict fine-grained RBox based on HBox annotations. Along with a shared backbone and neck structure, our proposed EIE-Det comprises a weakly supervised (WS) branch for predicting coarse RBox from HBox labels, an explicit equivariance (EE) branch for learning rotation consistency, and an implicit equivariance (IE) branch for learning position, aspect ratio, and scale consistency. Furthermore, we introduce the Tanimoto Coefficient for vector-based bounding box regression, enabling a comprehensive assessment of model prediction bias, thereby eliminating interference from variations in box aspect ratio and scale. Experiments demonstrate that our method not only significantly surpasses state-of-the-art models supervised with HBox but also achieves accuracy comparable to fully-supervised detectors (75.74$\%$ and 58.50$\%$ $AP_{50}$ on DOTA and DIOR-R, respectively).
- 3 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26 (15209210) , 5181-5193
Bounding Box Vectorization for Oriented Object Detection With Tanimoto Coefficient Regression
Current oriented object detection methods mainly utilize a vanilla coordinate-angle representation for bounding box regression, which usually suffers from inconsistency between the bounding box regression losses and prediction errors induced with respect to different rotation angles, aspect ratios, and scales. Therefore, although the existing oriented object detectors have achieved very good performances under coarse evaluation metrics such as AP50, their performance significantly degrades when using stricter evaluation metric such as AP75. To address the abovementioned issues, we propose a new regression method with bounding box vectorization that implicitly represents the shape and orientation of an object with a set of orthogonal vectors. By doing this, the proposed method delicately avoids the inconsistency issues encountered in oriented bounding box regression. During training, we introduce the Tanimoto coefficient to evaluate the similarity of the bounding box vector in a shape- and orientation-aware manner, and we refer to the proposed box-to-vector loss as the B2V loss. In addition to 2D object detection, the proposed method can be easily generalized to 3D scenarios involving orientation estimation, such as autonomous driving. We evaluate the proposed method through extensive experiments conducted on four popular oriented object detection datasets, including both 2D and 3D datasets, where the proposed method significantly outperforms the recently developed state-of-the-art methods when using a more accurate evaluation metric.
- 3 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26 (15209210) , 5181-5193
Bounding Box Vectorization for Oriented Object Detection With Tanimoto Coefficient Regression
Current oriented object detection methods mainly utilize a vanilla coordinate-angle representation for bounding box regression, which usually suffers from inconsistency between the bounding box regression losses and prediction errors induced with respect to different rotation angles, aspect ratios, and scales. Therefore, although the existing oriented object detectors have achieved very good performances under coarse evaluation metrics such as AP50, their performance significantly degrades when using stricter evaluation metric such as AP75. To address the abovementioned issues, we propose a new regression method with bounding box vectorization that implicitly represents the shape and orientation of an object with a set of orthogonal vectors. By doing this, the proposed method delicately avoids the inconsistency issues encountered in oriented bounding box regression. During training, we introduce the Tanimoto coefficient to evaluate the similarity of the bounding box vector in a shape- and orientation-aware manner, and we refer to the proposed box-to-vector loss as the B2V loss. In addition to 2D object detection, the proposed method can be easily generalized to 3D scenarios involving orientation estimation, such as autonomous driving. We evaluate the proposed method through extensive experiments conducted on four popular oriented object detection datasets, including both 2D and 3D datasets, where the proposed method significantly outperforms the recently developed state-of-the-art methods when using a more accurate evaluation metric.
- 1 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Detection Refinement using Geometric Features for Robust Robotic Grasping
End-to-end grasp detection networks show state-of-the-art grasp accuracy in the standard datasets; these datasets use oriented rectangles to represent grasp poses and evaluate based on intersection over union (IoU) score. However, the model trained on these datasets, while being able to accurately localize to the object itself, does not necessarily predict the robust grasp pose, as it does not take into account the geometric features of the object. In this paper, we propose a Grasp Refinement Algorithm using RGBD images, which constrains the grasping detection network to output more robust grasp poses by extracting shallow features (e.g., object central axis) of the target object. The grasp refinement algorithm mainly consists of a feature extraction module and a grasp refinement module. The feature extraction module extracts the central axis of the target object and the object skeleton based on the depth threshold, which are then fed sequentially into the grasp refinement module to optimize the grasp pose predicted by the grasp detection network. The proposed algorithm achieves state-of-the-art grasp detection accuracy on both the Cornell grasp dataset, the Jacquard grasp dataset, and a simulated grasp environment based on the YCB dataset, which shows that our algorithm outperforms its counterparts in terms of grasp robustness.
- 1 February 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Detection Refinement using Geometric Features for Robust Robotic Grasping
End-to-end grasp detection networks show state-of-the-art grasp accuracy in the standard datasets; these datasets use oriented rectangles to represent grasp poses and evaluate based on intersection over union (IoU) score. However, the model trained on these datasets, while being able to accurately localize to the object itself, does not necessarily predict the robust grasp pose, as it does not take into account the geometric features of the object. In this paper, we propose a Grasp Refinement Algorithm using RGBD images, which constrains the grasping detection network to output more robust grasp poses by extracting shallow features (e.g., object central axis) of the target object. The grasp refinement algorithm mainly consists of a feature extraction module and a grasp refinement module. The feature extraction module extracts the central axis of the target object and the object skeleton based on the depth threshold, which are then fed sequentially into the grasp refinement module to optimize the grasp pose predicted by the grasp detection network. The proposed algorithm achieves state-of-the-art grasp detection accuracy on both the Cornell grasp dataset, the Jacquard grasp dataset, and a simulated grasp environment based on the YCB dataset, which shows that our algorithm outperforms its counterparts in terms of grasp robustness.