Chaozhi Yang
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
Chaozhi Yang has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Zongmin Li
College of Computer Science and Technology, China University of Petroleum (East China), Qingdao, China
10 shared publications - Qian Xiao
China University of Petroleum (East China)
7 shared publications - Yachuan Li
China University of Petroleum (East China)
7 shared publications - Yun Bai
Qingdao Institute of Software, College of Computer Science and Technology, China University of Petroleum (East China), Qingdao, China
6 shared publications - Guanlin Li
College of Computer Science and Technology, China University of Petroleum (East China), Qingdao 266580, China
4 shared publications Qian Xiao
Qingdao Institute of Software, College of Computer Science and Technology, China University of Petroleum (East China), Qingdao, China
4 shared publications
Latest Publications
- 16 October 2025
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 11 (6) , 065014
Cerebrovascular segmentation network based on fast fourier convolution and Mamba
Purpose. Cerebrovascular segmentation is crucial for the diagnosis and treatment of cerebrovascular diseases. However, accurately extracting cerebral vessels from Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) remains challenging due to the topological complexity and anatomical variability. Methods. This paper presents a novel Y-shaped segmentation network with fast Fourier convolution and Mamba, termed F-Mamba-YNet. The network employs a dual-encoder architecture that effectively leverages the complementarity of spectral and spatial domains for achieving the fusion of multi-level features. The spectral encoder features the Fast Fourier Convolution Module, which captures high-frequency changes in vessel edges, improving segmentation completeness and connectivity. The spatial encoder incorporates a Spatial Mamba Module, which captures long-range dependencies while enhancing the spatial feature representation of cerebral vessels. Additionally, a Multi-scale Feature Selection Module in the decoder adaptively enhances discriminative features, enabling improved feature reuse. Results. Experiments demonstrate that the proposed F-Mamba-YNet achieved 86.28% and 72.24% Dice Similarity Coefficient (DSC) on the IXI-A-SegAN dataset and MIDAS dataset. Conclusions. Compared with existing algorithms, F-Mamba-YNet provided more connected and continuous segmentation results and achieved competitive performance in terms of generalization.
- 16 October 2025
- Published by IOP Publishing in Biomedical Physics & Engineering Express
- Vol. 11 (6) , 065014
Cerebrovascular segmentation network based on fast fourier convolution and Mamba
Purpose. Cerebrovascular segmentation is crucial for the diagnosis and treatment of cerebrovascular diseases. However, accurately extracting cerebral vessels from Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) remains challenging due to the topological complexity and anatomical variability. Methods. This paper presents a novel Y-shaped segmentation network with fast Fourier convolution and Mamba, termed F-Mamba-YNet. The network employs a dual-encoder architecture that effectively leverages the complementarity of spectral and spatial domains for achieving the fusion of multi-level features. The spectral encoder features the Fast Fourier Convolution Module, which captures high-frequency changes in vessel edges, improving segmentation completeness and connectivity. The spatial encoder incorporates a Spatial Mamba Module, which captures long-range dependencies while enhancing the spatial feature representation of cerebral vessels. Additionally, a Multi-scale Feature Selection Module in the decoder adaptively enhances discriminative features, enabling improved feature reuse. Results. Experiments demonstrate that the proposed F-Mamba-YNet achieved 86.28% and 72.24% Dice Similarity Coefficient (DSC) on the IXI-A-SegAN dataset and MIDAS dataset. Conclusions. Compared with existing algorithms, F-Mamba-YNet provided more connected and continuous segmentation results and achieved competitive performance in terms of generalization.
- 26 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 71 (4) , 9955-9967
Graph Convolutional Network for Point Cloud Understanding Through Homogeneity Partitioning
Effectively capturing complex point cloud information is essential for advanced functionalities in consumer electronics, such as augmented reality, virtual simulations, and 3D printing. Given the irregular and unstructured nature of point cloud data, graph-based methods have recently emerged as effective solutions. However, their uniform treatment of neighboring points often limits the ability to capture fine-grained local structural information, weakening discriminative features. To address this limitation, this paper introduces a graph convolutional network that incorporates homogeneity partitioning to enhance local structural modeling in point clouds. First, neighboring points are identified based on distance and connected to a central point to construct a local graph. Next, a graph vector is computed by summing unit edge vectors, capturing the dominant directional trend of the local neighborhood. Based on the similarity between edge vectors and this graph vector, the neighborhood is partitioned into two sets: a homogeneous point set, composed of points with consistent structural attributes, and a heterogeneous point set, composed of points with more diverse attributes. Then, a graph convolution operator is designed to capture node commonalities, suppress the influence of residual heterogeneous points within the homogeneous set, and extract shared information across different classes in the heterogeneous set. Finally, an interaction mechanism between the two sets is introduced to refine node features and incorporate richer contextual information. Extensive experiments on multiple datasets validate the effectiveness of the proposed method.
- 26 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 71 (4) , 9955-9967
Graph Convolutional Network for Point Cloud Understanding Through Homogeneity Partitioning
Effectively capturing complex point cloud information is essential for advanced functionalities in consumer electronics, such as augmented reality, virtual simulations, and 3D printing. Given the irregular and unstructured nature of point cloud data, graph-based methods have recently emerged as effective solutions. However, their uniform treatment of neighboring points often limits the ability to capture fine-grained local structural information, weakening discriminative features. To address this limitation, this paper introduces a graph convolutional network that incorporates homogeneity partitioning to enhance local structural modeling in point clouds. First, neighboring points are identified based on distance and connected to a central point to construct a local graph. Next, a graph vector is computed by summing unit edge vectors, capturing the dominant directional trend of the local neighborhood. Based on the similarity between edge vectors and this graph vector, the neighborhood is partitioned into two sets: a homogeneous point set, composed of points with consistent structural attributes, and a heterogeneous point set, composed of points with more diverse attributes. Then, a graph convolution operator is designed to capture node commonalities, suppress the influence of residual heterogeneous points within the homogeneous set, and extract shared information across different classes in the heterogeneous set. Finally, an interaction mechanism between the two sets is introduced to refine node features and incorporate richer contextual information. Extensive experiments on multiple datasets validate the effectiveness of the proposed method.
- 7 January 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 31 (1) , 1-15
Compact twice fusion network for edge detection
- 7 January 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 31 (1) , 1-15
Compact twice fusion network for edge detection
- 12 November 2024
- Published by Elsevier in Knowledge-Based Systems
- Vol. 306
Boosting point cloud understanding through graph convolutional network with scale measurement and high-frequency enhancement
- 12 November 2024
- Published by Elsevier in Knowledge-Based Systems
- Vol. 306
Boosting point cloud understanding through graph convolutional network with scale measurement and high-frequency enhancement
- 9 November 2024
- Published by Elsevier in Knowledge-Based Systems
- Vol. 306
ODDF-Net: Multi-object segmentation in 3D retinal OCTA using optical density and disease features
- 9 November 2024
- Published by Elsevier in Knowledge-Based Systems
- Vol. 306
ODDF-Net: Multi-object segmentation in 3D retinal OCTA using optical density and disease features
- 1 August 2024
- Published by Springer Nature
- p. 75-85
SF-MCTS: Score Feedback Monte Carlo Tree Search for Digital Curling in Continuous State Space
- 1 August 2024
- Published by Springer Nature
- p. 75-85
SF-MCTS: Score Feedback Monte Carlo Tree Search for Digital Curling in Continuous State Space
- 9 February 2024
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 69 (4) , 045019
LA-Net: layer attention network for 3D-to-2D retinal vessel segmentation in OCTA images
Objective. Retinal vessel segmentation from optical coherence tomography angiography (OCTA) volumes is significant in analyzing blood supply structures and the diagnosing ophthalmic diseases. However, accurate retinal vessel segmentation in 3D OCTA remains challenging due to the interference of choroidal blood flow signals and the variations in retinal vessel structure. Approach. This paper proposes a layer attention network (LA-Net) for 3D-to-2D retinal vessel segmentation. The network comprises a 3D projection path and a 2D segmentation path. The key component in the 3D path is the proposed multi-scale layer attention module, which effectively learns the layer features of OCT and OCTA to attend to the retinal vessel layer while suppressing the choroidal vessel layer. This module also efficiently captures 3D multi-scale information for improved semantic understanding during projection. In the 2D path, a reverse boundary attention module is introduced to explore and preserve boundary and shape features of retinal vessels by focusing on non-salient regions in deep features. Main results. Experimental results in two subsets of the OCTA-500 dataset showed that our method achieves advanced segmentation performance with Dice similarity coefficients of 93.04% and 89.74%, respectively. Significance. The proposed network provides reliable 3D-to-2D segmentation of retinal vessels, with potential for application in various segmentation tasks that involve projecting the input image. Implementation code: https://github.com/y8421036/LA-Net .
- 9 February 2024
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 69 (4) , 045019
LA-Net: layer attention network for 3D-to-2D retinal vessel segmentation in OCTA images
Objective. Retinal vessel segmentation from optical coherence tomography angiography (OCTA) volumes is significant in analyzing blood supply structures and the diagnosing ophthalmic diseases. However, accurate retinal vessel segmentation in 3D OCTA remains challenging due to the interference of choroidal blood flow signals and the variations in retinal vessel structure. Approach. This paper proposes a layer attention network (LA-Net) for 3D-to-2D retinal vessel segmentation. The network comprises a 3D projection path and a 2D segmentation path. The key component in the 3D path is the proposed multi-scale layer attention module, which effectively learns the layer features of OCT and OCTA to attend to the retinal vessel layer while suppressing the choroidal vessel layer. This module also efficiently captures 3D multi-scale information for improved semantic understanding during projection. In the 2D path, a reverse boundary attention module is introduced to explore and preserve boundary and shape features of retinal vessels by focusing on non-salient regions in deep features. Main results. Experimental results in two subsets of the OCTA-500 dataset showed that our method achieves advanced segmentation performance with Dice similarity coefficients of 93.04% and 89.74%, respectively. Significance. The proposed network provides reliable 3D-to-2D segmentation of retinal vessels, with potential for application in various segmentation tasks that involve projecting the input image. Implementation code: https://github.com/y8421036/LA-Net .
- 25 December 2023
- Published by Springer Nature
- p. 261-272
PiDiNeXt: An Efficient Edge Detector Based on Parallel Pixel Difference Networks
- 25 December 2023
- Published by Springer Nature
- p. 261-272
PiDiNeXt: An Efficient Edge Detector Based on Parallel Pixel Difference Networks
- 10 November 2023
- Published by Springer Nature
- p. 129-140
KDED: A Knowledge Distillation Based Edge Detector
- 10 November 2023
- Published by Springer Nature
- p. 129-140
KDED: A Knowledge Distillation Based Edge Detector
- Open Access
Compact Twice Fusion Network for Edge Detection
The significance of multi-scale features has been gradually recognized by the edge detection community. However, the fusion of multi-scale features increases the complexity of the model, which is not friendly to practical application. In this work, we propose a Compact Twice Fusion Network (CTFN) to fully integrate multi-scale features while maintaining the compactness of the model. CTFN includes two lightweight multi-scale feature fusion modules: a Semantic Enhancement Module (SEM) that can utilize the semantic information contained in coarse-scale features to guide the learning of fine-scale features, and a Pseudo Pixel-level Weighting (PPW) module that aggregate the complementary merits of multi-scale features by assigning weights to all features. Notwithstanding all this, the interference of texture noise makes the correct classification of some pixels still a challenge. For these hard samples, we propose a novel loss function, coined Dynamic Focal Loss, which reshapes the standard cross-entropy loss and dynamically adjusts the weights to correct the distribution of hard samples. We evaluate our method on three datasets, i.e., BSDS500, NYUDv2, and BIPEDv2. Compared with state-of-the-art methods, CTFN achieves competitive accuracy with less parameters and computational cost. Apart from the backbone, CTFN requires only 0.1M additional parameters, which reduces its computation cost to just 60% of other state-of-the-art methods. The codes are available at https://github.com/Li-yachuan/CTFN-pytorch-master.
Open AccessCompact Twice Fusion Network for Edge Detection
The significance of multi-scale features has been gradually recognized by the edge detection community. However, the fusion of multi-scale features increases the complexity of the model, which is not friendly to practical application. In this work, we propose a Compact Twice Fusion Network (CTFN) to fully integrate multi-scale features while maintaining the compactness of the model. CTFN includes two lightweight multi-scale feature fusion modules: a Semantic Enhancement Module (SEM) that can utilize the semantic information contained in coarse-scale features to guide the learning of fine-scale features, and a Pseudo Pixel-level Weighting (PPW) module that aggregate the complementary merits of multi-scale features by assigning weights to all features. Notwithstanding all this, the interference of texture noise makes the correct classification of some pixels still a challenge. For these hard samples, we propose a novel loss function, coined Dynamic Focal Loss, which reshapes the standard cross-entropy loss and dynamically adjusts the weights to correct the distribution of hard samples. We evaluate our method on three datasets, i.e., BSDS500, NYUDv2, and BIPEDv2. Compared with state-of-the-art methods, CTFN achieves competitive accuracy with less parameters and computational cost. Apart from the backbone, CTFN requires only 0.1M additional parameters, which reduces its computation cost to just 60% of other state-of-the-art methods. The codes are available at https://github.com/Li-yachuan/CTFN-pytorch-master.