Ming Zhao
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
Ming Zhao has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Boyang Chen
National Satellite Meteorological Center, No. 46, Zhongguancun South Street, Haidian District, Beijing 100081, China
1 shared publication - Cheng Gao
College of Hydrology and Water Resources, Hohai University, Nanjing 210024, China
1 shared publication Cheng Yan
Department of Information Engineering, Shanghai Maritime University, China
1 shared publicationFeng Xiao Tang
Central-South University
1 shared publicationGuixiang Zhang
College of Information Engineering, Shanghai Maritime University, Shanghai, People’s Republic of China
1 shared publicationHrishikesh Aradhye
Google Inc., 1600 Amphitheatre Parkway, Mountain View, CA 94043, USA
1 shared publication
Latest Publications
- 6 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (8) , 5762-5772
An Arbitrary-Scale Super-Resolution Network for Multi-Contrast MRI With Permuted Cross-Attention
In magnetic resonance imaging (MRI), low-resolution (LR) images often hamper clinical diagnosis and research due to constraints in imaging conditions and technology limitations. Recent studies in super-resolution (SR) reconstruction of multi-contrast MRI have shown promise by leveraging the complementary information from different MRI contrasts. However, existing multi-contrast MRI SR techniques face several challenges: 1) a lack of pre-alignment precision can result in distorted reconstructions; 2) prevailing transformer network structures, with their smaller windows (e.g., 8×8), struggle to effectively capture long-range dependencies and lack the ability to interact between different windows; and 3) current methods are limited to fixed integer scaling (e.g., 2×, 3×, 4×), which limits flexibility and increases complexity in training and storage. To address these challenges, we propose a novel arbitrary-scale SR network for multi-contrast MRI. Specifically, our approach compensates for spatial misalignment between modalities through deformable registration module and employs permuted cross-attention transformer in MR images. In addition, we introduce a ref-scale ensemble implicit attention module that better integrates high-frequency information from reference images and enables arbitrary-scale upsampling. Extensive experiments on two publicly available MRI datasets validate the superiority of our method in multi-contrast MRI SR, demonstrating its significant potential in clinical applications. Our code is available at https://github.com/fangxiaojia0/An-Arbitrary-Scale-Super-Resolution-Network-for-Multi-Contrast-MRI-With-Permuted-Cross-Attention.
- 6 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (8) , 5762-5772
An Arbitrary-Scale Super-Resolution Network for Multi-Contrast MRI With Permuted Cross-Attention
In magnetic resonance imaging (MRI), low-resolution (LR) images often hamper clinical diagnosis and research due to constraints in imaging conditions and technology limitations. Recent studies in super-resolution (SR) reconstruction of multi-contrast MRI have shown promise by leveraging the complementary information from different MRI contrasts. However, existing multi-contrast MRI SR techniques face several challenges: 1) a lack of pre-alignment precision can result in distorted reconstructions; 2) prevailing transformer network structures, with their smaller windows (e.g., 8×8), struggle to effectively capture long-range dependencies and lack the ability to interact between different windows; and 3) current methods are limited to fixed integer scaling (e.g., 2×, 3×, 4×), which limits flexibility and increases complexity in training and storage. To address these challenges, we propose a novel arbitrary-scale SR network for multi-contrast MRI. Specifically, our approach compensates for spatial misalignment between modalities through deformable registration module and employs permuted cross-attention transformer in MR images. In addition, we introduce a ref-scale ensemble implicit attention module that better integrates high-frequency information from reference images and enables arbitrary-scale upsampling. Extensive experiments on two publicly available MRI datasets validate the superiority of our method in multi-contrast MRI SR, demonstrating its significant potential in clinical applications. Our code is available at https://github.com/fangxiaojia0/An-Arbitrary-Scale-Super-Resolution-Network-for-Multi-Contrast-MRI-With-Permuted-Cross-Attention.
- Open Access
- 21 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 449-455
Physical Prior-based Underwater Image Compression
- 21 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 449-455
Open AccessPhysical Prior-based Underwater Image Compression
- Open Access
- 10 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 1006-1012
Scribble-Guided Weakly-Supervised Edge-Enhanced Saliency Detection with Centroid Optimization
- 10 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 1006-1012
Open AccessScribble-Guided Weakly-Supervised Edge-Enhanced Saliency Detection with Centroid Optimization
- Open Access
- 17 December 2024
- Published by Frontiers Media SA in Frontiers in Neurorobotics
- Vol. 18, 1489658
Edge-guided feature fusion network for RGB-T salient object detection
Introduction: RGB-T Salient Object Detection (SOD) aims to accurately segment salient regions in both visible light and thermal infrared images. However, many existing methods overlook the critical complementarity between these modalities, which can enhance detection accuracy.Methods: We propose the Edge-Guided Feature Fusion Network (EGFF-Net), which consists of cross-modal feature extraction, edge-guided feature fusion, and salience map prediction. Firstly, the cross-modal feature extraction module captures and aggregates united and intersecting information in each local region of RGB and thermal images. Then, the edge-guided feature fusion module enhances the edge features of salient regions, considering that edge information is very helpful in refining significant area details. Moreover, a layer-by-layer decoding structure integrates multi-level features and generates the prediction of salience maps.Results: We conduct extensive experiments on three benchmark datasets and compare EGFF-Net with state-of-the-art methods. Our approach achieves superior performance, demonstrating the effectiveness of the proposed modules in improving both detection accuracy and boundary refinement.Discussion: The results highlight the importance of integrating cross-modal information and edge-guided fusion in RGB-T SOD. Our method outperforms existing techniques and provides a robust framework for future developments in multi-modal saliency detection.
- 17 December 2024
- Published by Frontiers Media SA in Frontiers in Neurorobotics
- Vol. 18, 1489658
Open AccessEdge-guided feature fusion network for RGB-T salient object detection
Introduction: RGB-T Salient Object Detection (SOD) aims to accurately segment salient regions in both visible light and thermal infrared images. However, many existing methods overlook the critical complementarity between these modalities, which can enhance detection accuracy.Methods: We propose the Edge-Guided Feature Fusion Network (EGFF-Net), which consists of cross-modal feature extraction, edge-guided feature fusion, and salience map prediction. Firstly, the cross-modal feature extraction module captures and aggregates united and intersecting information in each local region of RGB and thermal images. Then, the edge-guided feature fusion module enhances the edge features of salient regions, considering that edge information is very helpful in refining significant area details. Moreover, a layer-by-layer decoding structure integrates multi-level features and generates the prediction of salience maps.Results: We conduct extensive experiments on three benchmark datasets and compare EGFF-Net with state-of-the-art methods. Our approach achieves superior performance, demonstrating the effectiveness of the proposed modules in improving both detection accuracy and boundary refinement.Discussion: The results highlight the importance of integrating cross-modal information and edge-guided fusion in RGB-T SOD. Our method outperforms existing techniques and provides a robust framework for future developments in multi-modal saliency detection.
- Open Access
- 18 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 953-958
Visualization modeling method for ship collision risk based on AIS data
Maritime collisions represent a significant threat to maritime safety, and the reasonable quantification of collision risks between vessels is of paramount importance for the prevention of potential incidents. The detection and analysis of maritime collision risks have significant theoretical and practical value. However, there is a relative paucity of research on the visualisation of these risks. This paper makes a contribution to the field by developing a vessel risk quantification model based on the Vessel Conflict Ranking Operator (VCRO) with the objective of identifying collision situations from vast amounts of raw AIS (Automatic Identification System) data. Moreover, to address the shortcomings of existing visualisation models in dynamic real-time environments, such as inadequate adaptation to rapidly changing maritime traffic conditions and the necessity for enhanced response speed and accuracy in real-time risk assessment and early warning systems, this study introduces a novel visualisation modelling method. This method is concerned with the construction of a risk flow diagram which illustrates the changes in vessel position that occur after long-term navigation in a specific area. It also captures the dynamics of close encounters between multiple vessels and elucidates the long-term variation patterns of collision risks among two or more vessels.
- 18 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 953-958
Open AccessVisualization modeling method for ship collision risk based on AIS data
Maritime collisions represent a significant threat to maritime safety, and the reasonable quantification of collision risks between vessels is of paramount importance for the prevention of potential incidents. The detection and analysis of maritime collision risks have significant theoretical and practical value. However, there is a relative paucity of research on the visualisation of these risks. This paper makes a contribution to the field by developing a vessel risk quantification model based on the Vessel Conflict Ranking Operator (VCRO) with the objective of identifying collision situations from vast amounts of raw AIS (Automatic Identification System) data. Moreover, to address the shortcomings of existing visualisation models in dynamic real-time environments, such as inadequate adaptation to rapidly changing maritime traffic conditions and the necessity for enhanced response speed and accuracy in real-time risk assessment and early warning systems, this study introduces a novel visualisation modelling method. This method is concerned with the construction of a risk flow diagram which illustrates the changes in vessel position that occur after long-term navigation in a specific area. It also captures the dynamics of close encounters between multiple vessels and elucidates the long-term variation patterns of collision risks among two or more vessels.
- 8 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 21 (1545598X) , 1-5
A Semi-Supervised Image Registration Framework Based on Multimodal Cross-Attention
Registration of multimodal image pairs is a fundamental task in many remote sensing applications. To achieve accurate and low-cost remote sensing image registration, we propose a semi-supervised image registration framework based on multimodal cross-attention, which consists of the encoder for feature extraction, multimodal cross-attention module, and detection/descriptor decoders. We adopt positional encoding for feature maps to enhance the features with spatial contexts, especially for remote sensing images with large geometrical deformations. To learn common features that are independent of modalities between multimodal images, we proposed a multimodal cross-attention module to extract cross-modal features, which helps the detectors to extract more reliable matching keypoints. The network is trained in a semi-supervised manner, which requires only a small dataset of incompletely labeled images. To learn reliable keypoints from image pairs with inconsistent intensity and geometrical deformations, we randomly establish different geometrical mappings for the multimodal image pairs during training and then enrich the keypoint labels by continuously adding reliable keypoints extracted by the detection decoder in each epoch. Experimental results show that the proposed method achieves more comprehensive and accurate registration than the state-of-the-art methods for multimodal remote sensing images.
- 8 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 21 (1545598X) , 1-5
A Semi-Supervised Image Registration Framework Based on Multimodal Cross-Attention
Registration of multimodal image pairs is a fundamental task in many remote sensing applications. To achieve accurate and low-cost remote sensing image registration, we propose a semi-supervised image registration framework based on multimodal cross-attention, which consists of the encoder for feature extraction, multimodal cross-attention module, and detection/descriptor decoders. We adopt positional encoding for feature maps to enhance the features with spatial contexts, especially for remote sensing images with large geometrical deformations. To learn common features that are independent of modalities between multimodal images, we proposed a multimodal cross-attention module to extract cross-modal features, which helps the detectors to extract more reliable matching keypoints. The network is trained in a semi-supervised manner, which requires only a small dataset of incompletely labeled images. To learn reliable keypoints from image pairs with inconsistent intensity and geometrical deformations, we randomly establish different geometrical mappings for the multimodal image pairs during training and then enrich the keypoint labels by continuously adding reliable keypoints extracted by the detection decoder in each epoch. Experimental results show that the proposed method achieves more comprehensive and accurate registration than the state-of-the-art methods for multimodal remote sensing images.
- 3 May 2024
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 15 (5) , 558-569
A class distribution learning method for few-shot remote sensing scene classification
1. Remote sensing image scene classification has achieved stunning success in various applications, such as urban planning, hyperspectral-X image classification and land-use classification (Wang, L...
- 3 May 2024
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 15 (5) , 558-569
A class distribution learning method for few-shot remote sensing scene classification
1. Remote sensing image scene classification has achieved stunning success in various applications, such as urban planning, hyperspectral-X image classification and land-use classification (Wang, L...
- 3 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 20 (1545598X) , 1-5
A Self-Adaptive Object Detection Network for Aerial Images Based on Feature Enhancement
Object detection in aerial images has received increasing attention for its widely applications. However, it is still a challenging task when dealing with difficult cases, such as the large variations of object sizes, the complex backgrounds in the large-scale views, as well as small targets packed in dense. In this regard, we propose a self-adaptive object detection network for aerial images based on feature enhancement, including region crop module with soft-attention (RCP), feature enhancement module (FET), self adaptive feature extraction module (SAE). Firstly, RCP module with soft-attention is explored to roughly crop the dense subregion and sparse subregion into patches accoridng to the variance of feature maps for the following feature extraction and detection. Secondly, FET module is proposed to acquire more semantic details by feature enhancement, which makes up the information loss during downsampling, especially for small objects in dense regions. Finally, SAE module is explored to effectively identify multiple target regions and single target regions in the dense patches. The similarity of adjacent dense areas in dense patches is calcluated, and the search range is gradually narrowed by continuously merging the areas with the largest similarity. The teacher-network and student-network are used to extract features for multiple target regions and single target regions, respectively. The proposed design improves the accuracy of real-time detection under the drone’s perspective. A large number of experiments and comprehensive evaluations on the VisDrone2019-DET dataset have shown the effectiveness and adaptability of the proposed method. Our source codes have been available at https://github.com/zhaokai152.
- 3 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 20 (1545598X) , 1-5
A Self-Adaptive Object Detection Network for Aerial Images Based on Feature Enhancement
Object detection in aerial images has received increasing attention for its widely applications. However, it is still a challenging task when dealing with difficult cases, such as the large variations of object sizes, the complex backgrounds in the large-scale views, as well as small targets packed in dense. In this regard, we propose a self-adaptive object detection network for aerial images based on feature enhancement, including region crop module with soft-attention (RCP), feature enhancement module (FET), self adaptive feature extraction module (SAE). Firstly, RCP module with soft-attention is explored to roughly crop the dense subregion and sparse subregion into patches accoridng to the variance of feature maps for the following feature extraction and detection. Secondly, FET module is proposed to acquire more semantic details by feature enhancement, which makes up the information loss during downsampling, especially for small objects in dense regions. Finally, SAE module is explored to effectively identify multiple target regions and single target regions in the dense patches. The similarity of adjacent dense areas in dense patches is calcluated, and the search range is gradually narrowed by continuously merging the areas with the largest similarity. The teacher-network and student-network are used to extract features for multiple target regions and single target regions, respectively. The proposed design improves the accuracy of real-time detection under the drone’s perspective. A large number of experiments and comprehensive evaluations on the VisDrone2019-DET dataset have shown the effectiveness and adaptability of the proposed method. Our source codes have been available at https://github.com/zhaokai152.
- 1 February 2022
- Published by Taylor & Francis in International Journal of Remote Sensing
- Vol. 43 (3) , 915-931
Heterogeneous self-supervised interest point matching for multi-modal remote sensing image registration
The implementation of multi-modal remote sensing image registration is still a challenge in various applications. In this paper, a novel framework called Heterogeneous SuperPoint Network is proposed for multi-modal remote sensing image registration. Heterogeneous SuperPoint Network operates on full-sized images and produces multi-modal interest points accompanied with fixed length descriptors in a single forward pass. Moreover, we develop a strategy for training a base detector that used in conjunction with heterogeneous homographic adaptation to generate pseudo-ground truth interest point labels for unlabeled images in a self-supervised fashion. Then, the interest points and point correspondences obtained by Heterogeneous SuperPoint Network are used to do the final matching with the nearest neighbor (NN) matching and Random sample consensus (RANSAC). Various image sets have been considered in this paper for the evaluation of the proposed approach. Experimental results demonstrate the superior performance of the proposed method over the existing design for the whole data set.
- 1 February 2022
- Published by Taylor & Francis in International Journal of Remote Sensing
- Vol. 43 (3) , 915-931
Heterogeneous self-supervised interest point matching for multi-modal remote sensing image registration
The implementation of multi-modal remote sensing image registration is still a challenge in various applications. In this paper, a novel framework called Heterogeneous SuperPoint Network is proposed for multi-modal remote sensing image registration. Heterogeneous SuperPoint Network operates on full-sized images and produces multi-modal interest points accompanied with fixed length descriptors in a single forward pass. Moreover, we develop a strategy for training a base detector that used in conjunction with heterogeneous homographic adaptation to generate pseudo-ground truth interest point labels for unlabeled images in a self-supervised fashion. Then, the interest points and point correspondences obtained by Heterogeneous SuperPoint Network are used to do the final matching with the nearest neighbor (NN) matching and Random sample consensus (RANSAC). Various image sets have been considered in this paper for the evaluation of the proposed approach. Experimental results demonstrate the superior performance of the proposed method over the existing design for the whole data set.
- 20 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Orientation-Aware Feature Fusion Network for Ship Detection in SAR Images
Recently, deep learning methods have been successfully applied to the ship detection in synthetic aperture radar (SAR) images. It is still a great challenge to detect SAR ships, due to the extremely poor image quality and complex background. To solve the problems, a novel method named orientation-aware feature fusion network (OFF-Net) for ship detection in SAR images is proposed in this letter. OFF-Net consists of global context path aggregation (GCPA) module and local rotated contrast enhance (LRCE) module, which fuses the global and local information in feature extraction. First, GCPA module is explored to integrate the global context block with path aggregation network (PAN) to learn the global background information. Second, by designing a rotation scheme based on feature map cyclic shift with four directions, LRCE module is developed to enhance the targets and suppress the background clutters in SAR images. Finally, a decoupled orientation-aware head is proposed to handle the arbitrarily rotated ships more robustly and alleviate the conflict between classification and regression tasks during detection. In addition, a high-resolution SAR-ship detection dataset (OBB-HRSDD) with rotatable bounding boxes is provided. The detection results on the SAR ship detection dataset (SSDD+) and OBB-HRSDD illustrate that our method outperforms all the compared methods. The code and OBB-HRSDD are released at https://github.com/SJX152/papercode
- 20 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Orientation-Aware Feature Fusion Network for Ship Detection in SAR Images
Recently, deep learning methods have been successfully applied to the ship detection in synthetic aperture radar (SAR) images. It is still a great challenge to detect SAR ships, due to the extremely poor image quality and complex background. To solve the problems, a novel method named orientation-aware feature fusion network (OFF-Net) for ship detection in SAR images is proposed in this letter. OFF-Net consists of global context path aggregation (GCPA) module and local rotated contrast enhance (LRCE) module, which fuses the global and local information in feature extraction. First, GCPA module is explored to integrate the global context block with path aggregation network (PAN) to learn the global background information. Second, by designing a rotation scheme based on feature map cyclic shift with four directions, LRCE module is developed to enhance the targets and suppress the background clutters in SAR images. Finally, a decoupled orientation-aware head is proposed to handle the arbitrarily rotated ships more robustly and alleviate the conflict between classification and regression tasks during detection. In addition, a high-resolution SAR-ship detection dataset (OBB-HRSDD) with rotatable bounding boxes is provided. The detection results on the SAR ship detection dataset (SSDD+) and OBB-HRSDD illustrate that our method outperforms all the compared methods. The code and OBB-HRSDD are released at https://github.com/SJX152/papercode