Ning Jing
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ning Jing has a total of 59 co-authors. Here are the top co-authors by number of shared publications:
- Luo Chen
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China
19 shared publications - Jun Li
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China
8 shared publications - Mengyu Ma
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China
8 shared publications - Anran Yang
Universal Intelligent Sensing Innovation Center, Tianjin Institute of Advanced Technology
6 shared publications - Hao Chen
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China
5 shared publications - Ye Wu
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, China
5 shared publications
Latest Publications
- Open Access
MscaVPR: Multi-Scale Coordinate Attention Network for Robust Visual Place Recognition
Visual place recognition (VPR) aims to localize a query image by matching its visual representation against a geotagged database. One major challenge in VPR is to learn place representations that remain robust under appearance changes, viewpoint variations, and perceptual aliasing. However, existing VPR methods still show limitations in adaptive multi-scale feature fusion and viewpoint-aware training supervision, which may hinder robustness under severe viewpoint changes. In this paper, we propose MscaVPR, a VPR framework that combines multi-scale feature enhancement with azimuth-aware training. Specifically, a Multi-Scale Spatial Pyramid Attention (MSPA) module is incorporated to aggregate regional features across different spatial scales, and Coordinate Attention (CA) is used to encode positional cues for spatially refined feature learning. To further enhance viewpoint robustness, we design an azimuth-guided training strategy that selects hard positive samples with significant viewpoint discrepancies and optimizes them using an azimuth-aware auxiliary loss function.Experimental results on multiple benchmark datasets indicate that MscaVPR generally outperforms the strong baseline and demonstrates improved performance under challenging conditions. In particular, Recall@1 is improved by 2.1%, 1.9%, and 1.9% on the AmsterTime, SVOX-Night, and SVOX-Sun datasets, respectively. The results demonstrate that explicitly incorporating azimuth cues provides an effective complement to existing multi-scale and attention-based VPR methods.
Open AccessMscaVPR: Multi-Scale Coordinate Attention Network for Robust Visual Place Recognition
Visual place recognition (VPR) aims to localize a query image by matching its visual representation against a geotagged database. One major challenge in VPR is to learn place representations that remain robust under appearance changes, viewpoint variations, and perceptual aliasing. However, existing VPR methods still show limitations in adaptive multi-scale feature fusion and viewpoint-aware training supervision, which may hinder robustness under severe viewpoint changes. In this paper, we propose MscaVPR, a VPR framework that combines multi-scale feature enhancement with azimuth-aware training. Specifically, a Multi-Scale Spatial Pyramid Attention (MSPA) module is incorporated to aggregate regional features across different spatial scales, and Coordinate Attention (CA) is used to encode positional cues for spatially refined feature learning. To further enhance viewpoint robustness, we design an azimuth-guided training strategy that selects hard positive samples with significant viewpoint discrepancies and optimizes them using an azimuth-aware auxiliary loss function.Experimental results on multiple benchmark datasets indicate that MscaVPR generally outperforms the strong baseline and demonstrates improved performance under challenging conditions. In particular, Recall@1 is improved by 2.1%, 1.9%, and 1.9% on the AmsterTime, SVOX-Night, and SVOX-Sun datasets, respectively. The results demonstrate that explicitly incorporating azimuth cues provides an effective complement to existing multi-scale and attention-based VPR methods.
- Open Access
- 12 January 2026
- Published by MDPI AG in Technologies
- Vol. 14 (1) , 58
HiFiRadio: High-Fidelity Radio Map Reconstruction for 3D Real-World Scenes
The reconstruction of high-fidelity radio maps is pivotal for wireless network planning but remains challenging due to the tension between physical accuracy and computational efficiency. We propose HiFiRadio, a novel framework that achieves a breakthrough in this balance by integrating centimeter-resolution 3D environmental meshes with semantic-aware propagation modeling. At its core, HiFiRadio introduces a semantic-enhanced 3D indexing structure that efficiently manages complex terrain data, enabling real-time classification of signal paths into line-of-sight, non-line-of-sight, and vegetation-obstructed categories. This classification directly guides a hybrid propagation model, which dynamically applies dedicated loss calculations for buildings and foliage, grounded in physical principles. Extensive experiments demonstrate that HiFiRadio attains an accuracy comparable to commercial ray-tracing tools while being orders of magnitude faster. It also significantly outperforms existing learning-based baselines in both accuracy and scalability, a claim further validated by field measurements. By making high-fidelity, real-time radio map reconstruction practical for large-scale scenes, HiFiRadio establishes a new state of the art with immediate applications in network planning, UAV pathing, and dynamic spectrum access.
- 12 January 2026
- Published by MDPI AG in Technologies
- Vol. 14 (1) , 58
Open AccessHiFiRadio: High-Fidelity Radio Map Reconstruction for 3D Real-World Scenes
The reconstruction of high-fidelity radio maps is pivotal for wireless network planning but remains challenging due to the tension between physical accuracy and computational efficiency. We propose HiFiRadio, a novel framework that achieves a breakthrough in this balance by integrating centimeter-resolution 3D environmental meshes with semantic-aware propagation modeling. At its core, HiFiRadio introduces a semantic-enhanced 3D indexing structure that efficiently manages complex terrain data, enabling real-time classification of signal paths into line-of-sight, non-line-of-sight, and vegetation-obstructed categories. This classification directly guides a hybrid propagation model, which dynamically applies dedicated loss calculations for buildings and foliage, grounded in physical principles. Extensive experiments demonstrate that HiFiRadio attains an accuracy comparable to commercial ray-tracing tools while being orders of magnitude faster. It also significantly outperforms existing learning-based baselines in both accuracy and scalability, a claim further validated by field measurements. By making high-fidelity, real-time radio map reconstruction practical for large-scale scenes, HiFiRadio establishes a new state of the art with immediate applications in network planning, UAV pathing, and dynamic spectrum access.
- Open Access
- 3 April 2025
- Published by MDPI AG in Journal of Marine Science and Engineering
- Vol. 13 (4) , 715
MESTR: A Multi-Task Enhanced Ship-Type Recognition Model Based on AIS
With the rapid growth in maritime traffic, navigational safety has become a pressing concern. Some vessels deliberately manipulate their type information to evade regulatory oversight, either to circumvent legal sanctions or engage in illicit activities. Such practices not only undermine the accuracy of maritime supervision but also pose significant risks to maritime traffic management and safety. Therefore, accurately identifying vessel types is essential for effective maritime traffic regulation, combating maritime crimes, and ensuring safe maritime transportation. However, the existing methods fail to fully exploit the long-term sequential dependencies and intricate mobility patterns embedded in vessel trajectory data, leading to suboptimal identification accuracy and reliability. To address these limitations, we propose MESTR, a Multi-Task Enhanced Ship-Type Recognition model based on Automatic Identification System (AIS) data. MESTR leverages a Transformer-based deep learning framework with a motion-pattern-aware trajectory segment masking strategy. By jointly optimizing two learning tasks—trajectory segment masking prediction and ship-type prediction—MESTR effectively captures deep spatiotemporal features of various vessel types. This approach enables the accurate classification of six common vessel categories: tug, sailing, fishing, passenger, tanker, and cargo. Experimental evaluations on real-world maritime datasets demonstrate the effectiveness of MESTR, achieving an average accuracy improvement of 12.04% over the existing methods.
- 3 April 2025
- Published by MDPI AG in Journal of Marine Science and Engineering
- Vol. 13 (4) , 715
Open AccessMESTR: A Multi-Task Enhanced Ship-Type Recognition Model Based on AIS
With the rapid growth in maritime traffic, navigational safety has become a pressing concern. Some vessels deliberately manipulate their type information to evade regulatory oversight, either to circumvent legal sanctions or engage in illicit activities. Such practices not only undermine the accuracy of maritime supervision but also pose significant risks to maritime traffic management and safety. Therefore, accurately identifying vessel types is essential for effective maritime traffic regulation, combating maritime crimes, and ensuring safe maritime transportation. However, the existing methods fail to fully exploit the long-term sequential dependencies and intricate mobility patterns embedded in vessel trajectory data, leading to suboptimal identification accuracy and reliability. To address these limitations, we propose MESTR, a Multi-Task Enhanced Ship-Type Recognition model based on Automatic Identification System (AIS) data. MESTR leverages a Transformer-based deep learning framework with a motion-pattern-aware trajectory segment masking strategy. By jointly optimizing two learning tasks—trajectory segment masking prediction and ship-type prediction—MESTR effectively captures deep spatiotemporal features of various vessel types. This approach enables the accurate classification of six common vessel categories: tug, sailing, fishing, passenger, tanker, and cargo. Experimental evaluations on real-world maritime datasets demonstrate the effectiveness of MESTR, achieving an average accuracy improvement of 12.04% over the existing methods.
- 3 November 2024
- Published by Springer Nature
- p. 66-80
SemVG: Semantic Fused Feature Extraction Network for Visual Geo-Localization Under Urban Street Scenes
- 3 November 2024
- Published by Springer Nature
- p. 66-80
SemVG: Semantic Fused Feature Extraction Network for Visual Geo-Localization Under Urban Street Scenes
- Open Access
- 31 May 2024
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 13 (6) , 184
Two-Stage Path Planning for Long-Distance Off-Road Path Planning Based on Terrain Data
In the face of increasing demands for tasks such as mountain rescue, geological exploration, and military operations in complex wilderness environments, planning an efficient walking route is crucial. To address the inefficiency of traditional two-dimensional path planning, this paper proposes a two-stage path planning algorithm. First, an improved Probabilistic Roadmap (PRM) algorithm is used to quickly and roughly determine the initial path. Then, the morphological dilation is applied to process the grid points of the initial path, retaining the surrounding area of the initial path for a precise positioning of the search range. Finally, the idea of the A∗ algorithm is applied to achieve precise path planning in the refined search range. During the process of constructing the topology map, we utilized parallelization acceleration strategies to expedite the graph construction. In order to verify the effectiveness of the algorithm, we used terrain data to construct a wilderness environment model, and tests were conducted on off-road path planning tasks with different terrains and distances. The experimental results show a substantial enhancement in the computational efficiency of the proposed algorithm relative to the conventional A∗ algorithm by 30 to 60 times.
- 31 May 2024
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 13 (6) , 184
Open AccessTwo-Stage Path Planning for Long-Distance Off-Road Path Planning Based on Terrain Data
In the face of increasing demands for tasks such as mountain rescue, geological exploration, and military operations in complex wilderness environments, planning an efficient walking route is crucial. To address the inefficiency of traditional two-dimensional path planning, this paper proposes a two-stage path planning algorithm. First, an improved Probabilistic Roadmap (PRM) algorithm is used to quickly and roughly determine the initial path. Then, the morphological dilation is applied to process the grid points of the initial path, retaining the surrounding area of the initial path for a precise positioning of the search range. Finally, the idea of the A∗ algorithm is applied to achieve precise path planning in the refined search range. During the process of constructing the topology map, we utilized parallelization acceleration strategies to expedite the graph construction. In order to verify the effectiveness of the algorithm, we used terrain data to construct a wilderness environment model, and tests were conducted on off-road path planning tasks with different terrains and distances. The experimental results show a substantial enhancement in the computational efficiency of the proposed algorithm relative to the conventional A∗ algorithm by 30 to 60 times.
- Open Access
- 14 February 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (4) , 678
GeoViewMatch: A Multi-Scale Feature-Matching Network for Cross-View Geo-Localization Using Swin-Transformer and Contrastive Learning
Cross-view geo-localization aims to locate street-view images by matching them with a collection of GPS-tagged remote sensing (RS) images. Due to the significant viewpoint and appearance differences between street-view images and RS images, this task is highly challenging. While deep learning-based methods have shown their dominance in the cross-view geo-localization task, existing models have difficulties in extracting comprehensive meaningful features from both domains of images. This limitation results in not establishing accurate and robust dependencies between street-view images and the corresponding RS images. To address the aforementioned issues, this paper proposes a novel and lightweight neural network for cross-view geo-localization. Firstly, in order to capture more diverse information, we propose a module for extracting multi-scale features from images. Secondly, we introduce contrastive learning and design a contrastive loss to further enhance the robustness in extracting and aligning meaningful multi-scale features. Finally, we conduct comprehensive experiments on two open benchmarks. The experimental results have demonstrated the superiority of the proposed method over the state-of-the-art methods.
- 14 February 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (4) , 678
Open AccessGeoViewMatch: A Multi-Scale Feature-Matching Network for Cross-View Geo-Localization Using Swin-Transformer and Contrastive Learning
Cross-view geo-localization aims to locate street-view images by matching them with a collection of GPS-tagged remote sensing (RS) images. Due to the significant viewpoint and appearance differences between street-view images and RS images, this task is highly challenging. While deep learning-based methods have shown their dominance in the cross-view geo-localization task, existing models have difficulties in extracting comprehensive meaningful features from both domains of images. This limitation results in not establishing accurate and robust dependencies between street-view images and the corresponding RS images. To address the aforementioned issues, this paper proposes a novel and lightweight neural network for cross-view geo-localization. Firstly, in order to capture more diverse information, we propose a module for extracting multi-scale features from images. Secondly, we introduce contrastive learning and design a contrastive loss to further enhance the robustness in extracting and aligning meaningful multi-scale features. Finally, we conduct comprehensive experiments on two open benchmarks. The experimental results have demonstrated the superiority of the proposed method over the state-of-the-art methods.
- Open Access
- 28 August 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (17) , 4215
RemainNet: Explore Road Extraction from Remote Sensing Image Using Mask Image Modeling
Road extraction from a remote sensing image is a research hotspot due to its broad range of applications. Despite recent advancements, achieving precise road extraction remains challenging. Since a road is thin and long, roadside objects and shadows cause occlusions, thus influencing the distinguishment of the road. Masked image modeling reconstructs masked areas from unmasked areas, which is similar to the process of inferring occluded roads from nonoccluded areas. Therefore, we believe that mask image modeling is beneficial for indicating occluded areas from other areas, thus alleviating the occlusion issue in remote sensing image road extraction. In this paper, we propose a remote sensing image road extraction network named RemainNet, which is based on mask image modeling. RemainNet consists of a backbone, image prediction module, and semantic prediction module. An image prediction module reconstructs a masked area RGB value from unmasked areas. Apart from reconstructing original remote sensing images, a semantic prediction module of RemainNet also extracts roads from masked images. Extensive experiments are carried out on the Massachusetts Roads dataset and DeepGlobe Road Extraction dataset; the proposed RemainNet improves 0.82–1.70% IoU compared with other state-of-the-art road extraction methods.
- 28 August 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (17) , 4215
Open AccessRemainNet: Explore Road Extraction from Remote Sensing Image Using Mask Image Modeling
Road extraction from a remote sensing image is a research hotspot due to its broad range of applications. Despite recent advancements, achieving precise road extraction remains challenging. Since a road is thin and long, roadside objects and shadows cause occlusions, thus influencing the distinguishment of the road. Masked image modeling reconstructs masked areas from unmasked areas, which is similar to the process of inferring occluded roads from nonoccluded areas. Therefore, we believe that mask image modeling is beneficial for indicating occluded areas from other areas, thus alleviating the occlusion issue in remote sensing image road extraction. In this paper, we propose a remote sensing image road extraction network named RemainNet, which is based on mask image modeling. RemainNet consists of a backbone, image prediction module, and semantic prediction module. An image prediction module reconstructs a masked area RGB value from unmasked areas. Apart from reconstructing original remote sensing images, a semantic prediction module of RemainNet also extracts roads from masked images. Extensive experiments are carried out on the Massachusetts Roads dataset and DeepGlobe Road Extraction dataset; the proposed RemainNet improves 0.82–1.70% IoU compared with other state-of-the-art road extraction methods.
- Open Access
- 21 May 2023
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 12 (5) , 207
STO2Vec: A Multiscale Spatio-Temporal Object Representation Method for Association Analysis
Spatio-temporal association analysis has attracted attention in various fields, such as urban computing and crime analysis. The proliferation of positioning technology and location-based services has facilitated the expansion of association analysis across spatio-temporal scales. However, existing methods inadequately consider the scale differences among spatio-temporal objects during analysis, leading to suboptimal precision in association analysis results. To remedy this issue, we propose a multiscale spatio-temporal object representation method, STO2Vec, for association analysis. This method comprises of two parts: graph construction and embedding. For graph construction, we introduce an adaptive hierarchical discretization method to distinguish the varying scales of local features. Then, we merge the embedding method for spatio-temporal objects with that for discrete units, establishing a heterogeneous graph. For embedding, to enhance embedding quality for homogeneous and heterogeneous data, we use biased sampling and unsupervised models to capture the association strengths between spatio-temporal objects. Empirical results using real-world open-source datasets show that STO2Vec outperforms other models, improving accuracy by 16.25% on average across diverse applications. Further case studies indicate STO2Vec effectively detects association relationships between spatio-temporal objects in a range of scenarios and is applicable to tasks such as moving object behavior pattern mining and trajectory semantic annotation.
- 21 May 2023
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 12 (5) , 207
Open AccessSTO2Vec: A Multiscale Spatio-Temporal Object Representation Method for Association Analysis
Spatio-temporal association analysis has attracted attention in various fields, such as urban computing and crime analysis. The proliferation of positioning technology and location-based services has facilitated the expansion of association analysis across spatio-temporal scales. However, existing methods inadequately consider the scale differences among spatio-temporal objects during analysis, leading to suboptimal precision in association analysis results. To remedy this issue, we propose a multiscale spatio-temporal object representation method, STO2Vec, for association analysis. This method comprises of two parts: graph construction and embedding. For graph construction, we introduce an adaptive hierarchical discretization method to distinguish the varying scales of local features. Then, we merge the embedding method for spatio-temporal objects with that for discrete units, establishing a heterogeneous graph. For embedding, to enhance embedding quality for homogeneous and heterogeneous data, we use biased sampling and unsupervised models to capture the association strengths between spatio-temporal objects. Empirical results using real-world open-source datasets show that STO2Vec outperforms other models, improving accuracy by 16.25% on average across diverse applications. Further case studies indicate STO2Vec effectively detects association relationships between spatio-temporal objects in a range of scenarios and is applicable to tasks such as moving object behavior pattern mining and trajectory semantic annotation.
- Open Access
- 20 October 2021
- Published by MDPI AG in Remote Sensing
- Vol. 13 (21) , 4201
SGA-Net: Self-Constructing Graph Attention Neural Network for Semantic Segmentation of Remote Sensing Images
Semantic segmentation of remote sensing images is always a critical and challenging task. Graph neural networks, which can capture global contextual representations, can exploit long-range pixel dependency, thereby improving semantic segmentation performance. In this paper, a novel self-constructing graph attention neural network is proposed for such a purpose. Firstly, ResNet50 was employed as backbone of a feature extraction network to acquire feature maps of remote sensing images. Secondly, pixel-wise dependency graphs were constructed from the feature maps of images, and a graph attention network is designed to extract the correlations of pixels of the remote sensing images. Thirdly, the channel linear attention mechanism obtained the channel dependency of images, further improving the prediction of semantic segmentation. Lastly, we conducted comprehensive experiments and found that the proposed model consistently outperformed state-of-the-art methods on two widely used remote sensing image datasets.
- 20 October 2021
- Published by MDPI AG in Remote Sensing
- Vol. 13 (21) , 4201
Open AccessSGA-Net: Self-Constructing Graph Attention Neural Network for Semantic Segmentation of Remote Sensing Images
Semantic segmentation of remote sensing images is always a critical and challenging task. Graph neural networks, which can capture global contextual representations, can exploit long-range pixel dependency, thereby improving semantic segmentation performance. In this paper, a novel self-constructing graph attention neural network is proposed for such a purpose. Firstly, ResNet50 was employed as backbone of a feature extraction network to acquire feature maps of remote sensing images. Secondly, pixel-wise dependency graphs were constructed from the feature maps of images, and a graph attention network is designed to extract the correlations of pixels of the remote sensing images. Thirdly, the channel linear attention mechanism obtained the channel dependency of images, further improving the prediction of semantic segmentation. Lastly, we conducted comprehensive experiments and found that the proposed model consistently outperformed state-of-the-art methods on two widely used remote sensing image datasets.
- Open Access
HiVision: Rapid Visualization of Large-Scale Spatial Vector Data
Rapid visualization of large-scale spatial vector data is a long-standing challenge in Geographic Information Science. In existing methods, the computation overheads grow rapidly with data volumes, leading to the incapability of providing real-time visualization for large-scale spatial vector data, even with parallel acceleration technologies. To fill the gap, we present HiVision, a display-driven visualization model for large-scale spatial vector data. Different from traditional data-driven methods, the computing units in HiVision are pixels rather than spatial objects to achieve real-time performance, and efficient spatial-index-based strategies are introduced to estimate the topological relationships between pixels and spatial objects. HiVision can maintain exceedingly good performance regardless of the data volume due to the stable pixel number for display. In addition, an optimized parallel computing architecture is proposed in HiVision to ensure the ability of real-time visualization. Experiments show that our approach outperforms traditional methods in rendering speed and visual effects while dealing with large-scale spatial vector data, and can provide interactive visualization of datasets with billion-scale points/segments/edges in real-time with flexible rendering styles. The HiVision code is open-sourced at https://github.com/MemoryMmy/HiVision with an online demonstration.
Open AccessHiVision: Rapid Visualization of Large-Scale Spatial Vector Data
Rapid visualization of large-scale spatial vector data is a long-standing challenge in Geographic Information Science. In existing methods, the computation overheads grow rapidly with data volumes, leading to the incapability of providing real-time visualization for large-scale spatial vector data, even with parallel acceleration technologies. To fill the gap, we present HiVision, a display-driven visualization model for large-scale spatial vector data. Different from traditional data-driven methods, the computing units in HiVision are pixels rather than spatial objects to achieve real-time performance, and efficient spatial-index-based strategies are introduced to estimate the topological relationships between pixels and spatial objects. HiVision can maintain exceedingly good performance regardless of the data volume due to the stable pixel number for display. In addition, an optimized parallel computing architecture is proposed in HiVision to ensure the ability of real-time visualization. Experiments show that our approach outperforms traditional methods in rendering speed and visual effects while dealing with large-scale spatial vector data, and can provide interactive visualization of datasets with billion-scale points/segments/edges in real-time with flexible rendering styles. The HiVision code is open-sourced at https://github.com/MemoryMmy/HiVision with an online demonstration.