Pengming Feng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pengming Feng has a total of 85 co-authors. Here are the top co-authors by number of shared publications:
- Jian Guan
Group of Intelligent Signal Processing, Harbin Engineering University, Harbin, China
13 shared publications - Guangjun He
Beijing Institute of Satellite Information Engineering, Beijing, China
8 shared publications - Wenwu Wang
School of Computer Science and Electronic Engineering, University of Surrey, Guildford GU2 7XH, UK
5 shared publications - Xueliang Zhang
Jiangsu Provincial Key Laboratory for Advanced Remote Sensing and Geographic Information Technology, Key Laboratory for Land Satellite Remote Sensing Applications of Ministry of Natural Resources, School of Geography and Ocean Science, Nanjing University, Nanjing, Jiangsu 210023, China
5 shared publications - Mingjie Xie
Beihang University, Beijing, China
4 shared publications Guangjun He
Beijing Institute of Satellite Information Engineering
3 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 25-29
Structured Correspondence Fusion for Image-Driven UAV Navigation
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 25-29
Open AccessStructured Correspondence Fusion for Image-Driven UAV Navigation
- 1 October 2026
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 160  (4) , 3505-3516
Physics and metadata-guided generative learning for underwater acoustic target recognition
Underwater acoustic target recognition (UATR) is challenged by data scarcity and long-tailed class distributions, where limited recordings provide incomplete coverage of vessel attributes and recording conditions. We propose a physics and metadata-guided generative learning (PMGL) framework comprising physics and metadata-guided acoustic generation (PMAG), physical semantic coverage expansion (PSCE), and physical semantic distribution rebalancing (PSDR). PMAG constructs structured physical semantic conditions from metadata and environmental priors to guide text-to-audio generation. PSCE expands intraclass condition coverage for few-shot UATR, while PSDR supplements underrepresented categories to construct a more balanced training distribution for long-tailed UATR. Experiments on the ShipsEar and DeepShip datasets demonstrate consistent improvements in recognition performance over the original training setting and a baseline based on a generative adversarial network (GAN), particularly under data-scarce and distribution-imbalanced conditions. The results indicate that the proposed PMGL framework effectively compensates for structural data deficiencies in underwater acoustic recognition.
- 1 October 2026
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 160  (4) , 3505-3516
Physics and metadata-guided generative learning for underwater acoustic target recognition
Underwater acoustic target recognition (UATR) is challenged by data scarcity and long-tailed class distributions, where limited recordings provide incomplete coverage of vessel attributes and recording conditions. We propose a physics and metadata-guided generative learning (PMGL) framework comprising physics and metadata-guided acoustic generation (PMAG), physical semantic coverage expansion (PSCE), and physical semantic distribution rebalancing (PSDR). PMAG constructs structured physical semantic conditions from metadata and environmental priors to guide text-to-audio generation. PSCE expands intraclass condition coverage for few-shot UATR, while PSDR supplements underrepresented categories to construct a more balanced training distribution for long-tailed UATR. Experiments on the ShipsEar and DeepShip datasets demonstrate consistent improvements in recognition performance over the original training setting and a baseline based on a generative adversarial network (GAN), particularly under data-scarce and distribution-imbalanced conditions. The results indicate that the proposed PMGL framework effectively compensates for structural data deficiencies in underwater acoustic recognition.
- 11 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 64  (01962892) , 5627518
Frequency–Phase Aware Mamba Fusion for Multiscale Spatial Representation in Remote Sensing Image Interpretation
Remote sensing images (RSIs) exhibit complex spatial patterns due to diverse object scales, orientations, and distributions. These characteristics pose significant challenges for accurate interpretation. While existing methods focus on spatial feature extraction or rotational constraints, they often overlook the intrinsic relationship between frequency-phase information and spatial structures, which is critical for precise localization and classification. In this paper, we propose a generic and flexible feature fusion framework, called Phase-aware Mamba Fusion Mechanism (PMFM), which jointly captures frequency-phase patterns and global spatial dependencies. PMFM consists of two core modules, i.e., Frequency-Phase Position Encoding (PhasePE), which enhances positional encoding by embedding frequency-phase cues, and Cross-attention–Mamba Progressive Fusion (CaMPF), which integrates global-local information across multi-scale feature maps using cross-attention and Vision Mamba. Furthermore, we design a PMFM-based feature pyramid architecture (PMFM-FPN) via a cross-scale concatenation strategy, enabling seamless integration into various detection and segmentation frameworks. Extensive experiments on five public RSI benchmarks (DOTA, DIOR-R, SISP, ISPRS Potsdam, and Vaihingen) across three major tasks, i.e., horizontal detection, oriented detection, and semantic segmentation, demonstrate that PMFM consistently improves performance and achieves competitive accuracy and robustness. In addition, an auxiliary experiment on ExDark further explores the behavior of phase-aware representations under general low-light visual conditions.
- 11 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 64  (01962892) , 5627518
Frequency–Phase Aware Mamba Fusion for Multiscale Spatial Representation in Remote Sensing Image Interpretation
Remote sensing images (RSIs) exhibit complex spatial patterns due to diverse object scales, orientations, and distributions. These characteristics pose significant challenges for accurate interpretation. While existing methods focus on spatial feature extraction or rotational constraints, they often overlook the intrinsic relationship between frequency-phase information and spatial structures, which is critical for precise localization and classification. In this paper, we propose a generic and flexible feature fusion framework, called Phase-aware Mamba Fusion Mechanism (PMFM), which jointly captures frequency-phase patterns and global spatial dependencies. PMFM consists of two core modules, i.e., Frequency-Phase Position Encoding (PhasePE), which enhances positional encoding by embedding frequency-phase cues, and Cross-attention–Mamba Progressive Fusion (CaMPF), which integrates global-local information across multi-scale feature maps using cross-attention and Vision Mamba. Furthermore, we design a PMFM-based feature pyramid architecture (PMFM-FPN) via a cross-scale concatenation strategy, enabling seamless integration into various detection and segmentation frameworks. Extensive experiments on five public RSI benchmarks (DOTA, DIOR-R, SISP, ISPRS Potsdam, and Vaihingen) across three major tasks, i.e., horizontal detection, oriented detection, and semantic segmentation, demonstrate that PMFM consistently improves performance and achieves competitive accuracy and robustness. In addition, an auxiliary experiment on ExDark further explores the behavior of phase-aware representations under general low-light visual conditions.
- Open Access
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19  (19391404) , 9819-9833
Decoupling Static Context and Dynamic Change: A Spatiotemporal Decoder With Time-Averaged Priors for Multidecadal Heritage Monitoring
The long-term preservation of Urban Cultural Heritage (UCH), a key directive of the United Nations' Sustainable Development Goal (SDG) 11.4, requires robust monitoring of multi-decadal landscape dynamics driven by urbanization. However, analyzing the necessary long-term satellite archives reveals a significant spatio-temporal challenge. This multi-decade data presents two interwoven problems: first, complex, non-stationary temporal dynamics that require processing the entire sequence history to distinguish evolving characteristics; and second, a severe spatial ambiguity in recent imagery, where heritage sites become spectrally indistinguishable from the surrounding, dynamically evolved urban sprawl. This dual challenge renders conventional methods, which are often purely spatial or purely temporal, ineffective at resolving this ambiguity. To address this, we first introduce a novel, multi-decadal (1984-2024) spatio-temporal dataset for Xi'an, China, specifically curated to study this problem. We then propose a novel architectural framework designed to solve this conundrum by decoupling static spatial context from dynamic temporal evolution. Our primary contribution is a new decoding strategy that directly confronts the spatial ambiguity. Instead of sourcing skip connections from a single, ambiguous, recent frame, our decoder is fed a time-averaged composite derived from the entire sequence. This composite acts as a robust spatial prior that inherently enhances the separability between stable heritage sites and their evolved urban surroundings. Evaluated on our dataset, the proposed framework achieves state-of-the-art performance (leading by 2.6%-10.7% in mIoU) and yields a 3.2% mIoU gain attributable to the static-prior decoder in ablation studies, contributing significant improvement independent of the specific temporal model used. This work introduces a validated paradigm for robust, UCH monitoring. Our codes will be available at https://github.com/xian1234/ST-Mamba
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19  (19391404) , 9819-9833
Open AccessDecoupling Static Context and Dynamic Change: A Spatiotemporal Decoder With Time-Averaged Priors for Multidecadal Heritage Monitoring
The long-term preservation of Urban Cultural Heritage (UCH), a key directive of the United Nations' Sustainable Development Goal (SDG) 11.4, requires robust monitoring of multi-decadal landscape dynamics driven by urbanization. However, analyzing the necessary long-term satellite archives reveals a significant spatio-temporal challenge. This multi-decade data presents two interwoven problems: first, complex, non-stationary temporal dynamics that require processing the entire sequence history to distinguish evolving characteristics; and second, a severe spatial ambiguity in recent imagery, where heritage sites become spectrally indistinguishable from the surrounding, dynamically evolved urban sprawl. This dual challenge renders conventional methods, which are often purely spatial or purely temporal, ineffective at resolving this ambiguity. To address this, we first introduce a novel, multi-decadal (1984-2024) spatio-temporal dataset for Xi'an, China, specifically curated to study this problem. We then propose a novel architectural framework designed to solve this conundrum by decoupling static spatial context from dynamic temporal evolution. Our primary contribution is a new decoding strategy that directly confronts the spatial ambiguity. Instead of sourcing skip connections from a single, ambiguous, recent frame, our decoder is fed a time-averaged composite derived from the entire sequence. This composite acts as a robust spatial prior that inherently enhances the separability between stable heritage sites and their evolved urban surroundings. Evaluated on our dataset, the proposed framework achieves state-of-the-art performance (leading by 2.6%-10.7% in mIoU) and yields a 3.2% mIoU gain attributable to the static-prior decoder in ablation studies, contributing significant improvement independent of the specific temporal model used. This work introduces a validated paradigm for robust, UCH monitoring. Our codes will be available at https://github.com/xian1234/ST-Mamba
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 29666-29682
SRC-Net: Saliency-Guided Reconstruction and Cholesky Regression Network for Oriented SAR Object Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 29666-29682
Open AccessSRC-Net: Saliency-Guided Reconstruction and Cholesky Regression Network for Oriented SAR Object Detection
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 324-332
Edge-aware Affinity Enhancement for Image Manipulation Localization
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 324-332
Open AccessEdge-aware Affinity Enhancement for Image Manipulation Localization
- Open Access
- 16 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18  (19391404) , 15394-15411
HMS-Net: A Hierarchical Multilabel Fine-Grained Ship Detection Network in Remote Sensing Images
The advent of deep learning has significantly advanced the efficiency and accuracy of fine-grained ship detection in remote sensing images. However, existing methods often rely on flattened label supervision during training, neglecting critical inter-class relationships and hierarchical semantic information, which exacerbates the challenges of fine-grained ship detection. To address these limitations, this paper proposes HMS-Net, a novel framework that integrates hierarchical multi-label classification with fine-grained ship detection. HMS-Net incorporates a parallel multi-level semantic feature extraction and fusion network, which leverages the Multi-scale Region Feature Reextraction (MRFR) module to perform effective feature reextraction at different semantic levels. Additionally, a top-down feature fusion mechanism, combined with the Feature Guidance (FG) module, facilitates efficient knowledge transfer from upper-level to lower-level features. Furthermore, we design a hierarchical relationship constraint loss function to explicitly model inter-class relationships among ships, optimizing feature distance measurements between different classes. Extensive ablation and comparative experiments on the HRSC2016 and ShipRSImageNet datasets demonstrate the superiority of HMSNet. Moreover, our method has been successfully applied to various two-stage oriented object detection frameworks, consistently achieving performance improvements, thereby highlighting its plug-and-play versatility.
- 16 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18  (19391404) , 15394-15411
Open AccessHMS-Net: A Hierarchical Multilabel Fine-Grained Ship Detection Network in Remote Sensing Images
The advent of deep learning has significantly advanced the efficiency and accuracy of fine-grained ship detection in remote sensing images. However, existing methods often rely on flattened label supervision during training, neglecting critical inter-class relationships and hierarchical semantic information, which exacerbates the challenges of fine-grained ship detection. To address these limitations, this paper proposes HMS-Net, a novel framework that integrates hierarchical multi-label classification with fine-grained ship detection. HMS-Net incorporates a parallel multi-level semantic feature extraction and fusion network, which leverages the Multi-scale Region Feature Reextraction (MRFR) module to perform effective feature reextraction at different semantic levels. Additionally, a top-down feature fusion mechanism, combined with the Feature Guidance (FG) module, facilitates efficient knowledge transfer from upper-level to lower-level features. Furthermore, we design a hierarchical relationship constraint loss function to explicitly model inter-class relationships among ships, optimizing feature distance measurements between different classes. Extensive ablation and comparative experiments on the HRSC2016 and ShipRSImageNet datasets demonstrate the superiority of HMSNet. Moreover, our method has been successfully applied to various two-stage oriented object detection frameworks, consistently achieving performance improvements, thereby highlighting its plug-and-play versatility.
- 19 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 22  (1545598X) , 1-5
Airplane State Discrimination From Single-Temporal High-Resolution Remote Sensing Images
The lack of temporal information in single-temporal satellite remote sensing images presents a substantial challenge for target state discrimination. In this letter, a pioneering Remote Sensing Airplane State Discrimination Network (RSASDNet) is introduced, by leveraging the relationship between targets and their backgrounds in single-temporal high-resolution remote sensing images. To facilitate the study, we take airplane state discrimination as an example, and a Remote Sensing Airport Panoptic Segmentation with Airplane States Dataset (RSAPS-ASD) is constructed. Subsequently, a scene knowledge graph generation module is designed to establish scene knowledge representation by capturing spatial relationships between airplane instances and their surrounding environment. Furthermore, a novel graph-image hybrid convolution discrimination module is proposed to extract structural knowledge and spatial semantic information separately, ultimately leading to the performance improvement of target state discrimination. The effectiveness of the proposed method is validated using RSAPS-ASD, with experimental results demonstrating that RSASDNet achieves an impressive accuracy of 73.95% in airplane state discrimination
- 19 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 22  (1545598X) , 1-5
Airplane State Discrimination From Single-Temporal High-Resolution Remote Sensing Images
The lack of temporal information in single-temporal satellite remote sensing images presents a substantial challenge for target state discrimination. In this letter, a pioneering Remote Sensing Airplane State Discrimination Network (RSASDNet) is introduced, by leveraging the relationship between targets and their backgrounds in single-temporal high-resolution remote sensing images. To facilitate the study, we take airplane state discrimination as an example, and a Remote Sensing Airport Panoptic Segmentation with Airplane States Dataset (RSAPS-ASD) is constructed. Subsequently, a scene knowledge graph generation module is designed to establish scene knowledge representation by capturing spatial relationships between airplane instances and their surrounding environment. Furthermore, a novel graph-image hybrid convolution discrimination module is proposed to extract structural knowledge and spatial semantic information separately, ultimately leading to the performance improvement of target state discrimination. The effectiveness of the proposed method is validated using RSAPS-ASD, with experimental results demonstrating that RSASDNet achieves an impressive accuracy of 73.95% in airplane state discrimination
- 22 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26  (15209210) , 8476-8487
Integration of Global and Local Knowledge for Foreground Enhancing in Weakly Supervised Temporal Action Localization
Weakly Supervised Temporal Action Localization (WTAL) aims to identify the temporal duration of actions and classify the action categories with only video-level labels in the training stage. Motivated by the intuition that the attention maps generated from various views will assist in enhancing the foreground action temporal segments, in this paper we propose a WTAL pipeline based on a novel attention mechanism that effectively integrates global and local knowledge. Our attention mechanism is mainly composed of a global attention branch and a local attention branch. Specifically, the global attention branch is built on the inter-segment similarity to sparsely mine out the correlation knowledge within the entire video, while the local attention branch is built on the convolutional structure to densely aggregate the information within the fixed local respective field. Experiments on THUMOS14 and ActivityNet v1.3 datasets demonstrate the effectiveness of our proposed WTAL pipeline compared to state-of-the-art methods.
- 22 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26  (15209210) , 8476-8487
Integration of Global and Local Knowledge for Foreground Enhancing in Weakly Supervised Temporal Action Localization
Weakly Supervised Temporal Action Localization (WTAL) aims to identify the temporal duration of actions and classify the action categories with only video-level labels in the training stage. Motivated by the intuition that the attention maps generated from various views will assist in enhancing the foreground action temporal segments, in this paper we propose a WTAL pipeline based on a novel attention mechanism that effectively integrates global and local knowledge. Our attention mechanism is mainly composed of a global attention branch and a local attention branch. Specifically, the global attention branch is built on the inter-segment similarity to sparsely mine out the correlation knowledge within the entire video, while the local attention branch is built on the convolutional structure to densely aggregate the information within the fixed local respective field. Experiments on THUMOS14 and ActivityNet v1.3 datasets demonstrate the effectiveness of our proposed WTAL pipeline compared to state-of-the-art methods.
- Open Access
SISP: A Benchmark Dataset for Fine-grained Ship Instance Segmentation in Panchromatic Satellite Images
Fine-grained ship instance segmentation in satellite images holds considerable significance for monitoring maritime activities at sea. However, existing datasets often suffer from the scarcity of fine-grained information or pixel-wise localization annotations, as well as the insufficient image diversity and variations, thus limiting the research of this task. To this end, we propose a benchmark dataset for fine-grained Ship Instance Segmentation in Panchromatic satellite images, namely SISP, which contains 56,693 well-annotated ship instances with four fine-grained categories across 10,000 sliced images, and all the images are collected from SuperView-1 satellite with the resolution of 0.5m. Targets in the proposed SISP dataset have characteristics that are consistent with real satellite scenes, such as high class imbalance, various scenes, large variations in target densities and scales, and high inter-class similarity and intra-class diversity, all of which make the SISP dataset more suitable for real-world applications. In addition, we introduce a Dynamic Feature Refinement-assist Instance segmentation network, namely DFRInst, as the benchmark method for ship instance segmentation in satellite images, which can fortify the explicit representation of crucial features, thus improving the performance of ship instance segmentation. Experiments and analysis are performed on the proposed SISP dataset to evaluate the benchmark method and several state-of-the-art methods to establish baselines for facilitating future research. The proposed dataset and source codes will be available at: https://github.com/Justlovesmile/SISP.
Open AccessSISP: A Benchmark Dataset for Fine-grained Ship Instance Segmentation in Panchromatic Satellite Images
Fine-grained ship instance segmentation in satellite images holds considerable significance for monitoring maritime activities at sea. However, existing datasets often suffer from the scarcity of fine-grained information or pixel-wise localization annotations, as well as the insufficient image diversity and variations, thus limiting the research of this task. To this end, we propose a benchmark dataset for fine-grained Ship Instance Segmentation in Panchromatic satellite images, namely SISP, which contains 56,693 well-annotated ship instances with four fine-grained categories across 10,000 sliced images, and all the images are collected from SuperView-1 satellite with the resolution of 0.5m. Targets in the proposed SISP dataset have characteristics that are consistent with real satellite scenes, such as high class imbalance, various scenes, large variations in target densities and scales, and high inter-class similarity and intra-class diversity, all of which make the SISP dataset more suitable for real-world applications. In addition, we introduce a Dynamic Feature Refinement-assist Instance segmentation network, namely DFRInst, as the benchmark method for ship instance segmentation in satellite images, which can fortify the explicit representation of crucial features, thus improving the performance of ship instance segmentation. Experiments and analysis are performed on the proposed SISP dataset to evaluate the benchmark method and several state-of-the-art methods to establish baselines for facilitating future research. The proposed dataset and source codes will be available at: https://github.com/Justlovesmile/SISP.