Yu Tao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yu Tao has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Ruopeng Yang
Department of Information and Communication Command, PLA Information Support Force Engineering University, Wuhan 430000, China
7 shared publications - Yongqi Shi
Graduate school, National University of Defense Technology, Changsha 410073, China
7 shared publications - Bo Huang
Department of Information and Communication Command, PLA Information Support Force Engineering University, Wuhan 430000, China
6 shared publications - Yiwei Lu
Department of Information and Communication Command, Information Support Force Engineering University, Wuhan 430035, China
6 shared publications - Yihao Zhong
School of Information Communication, National University of Defense Technology, Changsha 410073, China
5 shared publications - Changsheng Yin
Department of Information and Communications Command, Information Support Force Engineering University, Wuhan 430019, China
4 shared publications
Latest Publications
- Open Access
Dynamic Task-Chain Reconfiguration for Cooperative Counter-UAV Defense via Multi-Agent LLM-Based Heuristic Design
The growing affordability, autonomy, and swarming of small unmanned aerial vehicles (UAVs) turn low-altitude defense from single-shot interception into a multi-node cooperative decision problem, in which the loss of sensing, coordination, or engagement nodes breaks the closed loops linking them. This study formulates their recovery as the dynamic reconfiguration of cooperative counter-UAV task chains. Given a pre-disturbance plan and a set of failed defending nodes, reconfiguration is modeled as a constrained bi-objective optimization balancing recovered engagement effectiveness against the change to the baseline plan and is solved by Multi-Agent Heuristic Evolution (MAHE), an automated heuristic design framework whose evolution, coordinator, repair, and reflection agents—driven by a large language model—evolve scoring heuristics for a fixed reconfiguration solver. Across instances of varying scale and under light-to-heavy node loss conditions, MAHE outperforms both a single-agent heuristic design counterpart and a range of hand-crafted solvers: on ten held-out test instances spanning 8–320 targets it attains the highest overall normalized hypervolume (0.947, versus 0.935 for the single-agent counterpart and 0.30–0.45 for the hand-crafted solvers) and the best mean rank (1.43 of six methods, ); the hand-crafted solvers lose most of their solution quality as the problem grows, whereas MAHE preserves it and sustains high recovery at a nearly constant reconfiguration cost. An ablation confirms that its agents contribute complementary gains. These simulation results indicate that automatically generated, reconfiguration-specific heuristics offer a scalable algorithmic foundation for dynamic, heterogeneous, and constraint-intensive counter-UAV task-chain reconfiguration.Open AccessDynamic Task-Chain Reconfiguration for Cooperative Counter-UAV Defense via Multi-Agent LLM-Based Heuristic Design
The growing affordability, autonomy, and swarming of small unmanned aerial vehicles (UAVs) turn low-altitude defense from single-shot interception into a multi-node cooperative decision problem, in which the loss of sensing, coordination, or engagement nodes breaks the closed loops linking them. This study formulates their recovery as the dynamic reconfiguration of cooperative counter-UAV task chains. Given a pre-disturbance plan and a set of failed defending nodes, reconfiguration is modeled as a constrained bi-objective optimization balancing recovered engagement effectiveness against the change to the baseline plan and is solved by Multi-Agent Heuristic Evolution (MAHE), an automated heuristic design framework whose evolution, coordinator, repair, and reflection agents—driven by a large language model—evolve scoring heuristics for a fixed reconfiguration solver. Across instances of varying scale and under light-to-heavy node loss conditions, MAHE outperforms both a single-agent heuristic design counterpart and a range of hand-crafted solvers: on ten held-out test instances spanning 8–320 targets it attains the highest overall normalized hypervolume (0.947, versus 0.935 for the single-agent counterpart and 0.30–0.45 for the hand-crafted solvers) and the best mean rank (1.43 of six methods, ); the hand-crafted solvers lose most of their solution quality as the problem grows, whereas MAHE preserves it and sustains high recovery at a nearly constant reconfiguration cost. An ablation confirms that its agents contribute complementary gains. These simulation results indicate that automatically generated, reconfiguration-specific heuristics offer a scalable algorithmic foundation for dynamic, heterogeneous, and constraint-intensive counter-UAV task-chain reconfiguration. - Open Access
- 17 June 2026
- Published by MDPI AG in Preprints.org
Dynamic Task-Chain Reconfiguration for Cooperative Counter-UAV Defense: A Multi-Agent Large Language Model Framework for Automated Heuristic Design
The growing affordability, autonomy, and swarming of small unmanned aerial vehicles (UAVs) turn low-altitude defense from single-shot interception into a multi-node cooperative decision problem, in which the loss of sensing, coordination, or engagement nodes breaks the closed loops linking them. This study formulates their recovery as the dynamic reconfiguration of cooperative counter-UAV task chains. Given a pre-disturbance plan and a set of failed defending nodes, reconfiguration is modeled as a constrained bi-objective optimization balancing recovered engagement effectiveness against the change to the baseline plan, and is solved by Multi-Agent Heuristic Evolution (MAHE), an automated heuristic design framework whose evolution, coordinator, repair, and reflection agents—driven by a large language model—evolve scoring heuristics for a fixed reconfiguration solver. Across instances of varying scale and under light-to-heavy node loss, MAHE outperforms both a single-agent heuristic-design counterpart and a range of hand-crafted solvers: the latter lose most of their solution quality as the problem grows, whereas MAHE preserves it and sustains high recovery at a nearly constant reconfiguration cost; an ablation confirms that its agents contribute complementary gains. These results indicate that automatically generated, reconfiguration-specific heuristics provide a scalable route to dynamic, heterogeneous, and constraint-intensive counter-UAV task planning.
- 17 June 2026
- Published by MDPI AG in Preprints.org
Open AccessDynamic Task-Chain Reconfiguration for Cooperative Counter-UAV Defense: A Multi-Agent Large Language Model Framework for Automated Heuristic Design
The growing affordability, autonomy, and swarming of small unmanned aerial vehicles (UAVs) turn low-altitude defense from single-shot interception into a multi-node cooperative decision problem, in which the loss of sensing, coordination, or engagement nodes breaks the closed loops linking them. This study formulates their recovery as the dynamic reconfiguration of cooperative counter-UAV task chains. Given a pre-disturbance plan and a set of failed defending nodes, reconfiguration is modeled as a constrained bi-objective optimization balancing recovered engagement effectiveness against the change to the baseline plan, and is solved by Multi-Agent Heuristic Evolution (MAHE), an automated heuristic design framework whose evolution, coordinator, repair, and reflection agents—driven by a large language model—evolve scoring heuristics for a fixed reconfiguration solver. Across instances of varying scale and under light-to-heavy node loss, MAHE outperforms both a single-agent heuristic-design counterpart and a range of hand-crafted solvers: the latter lose most of their solution quality as the problem grows, whereas MAHE preserves it and sustains high recovery at a nearly constant reconfiguration cost; an ablation confirms that its agents contribute complementary gains. These results indicate that automatically generated, reconfiguration-specific heuristics provide a scalable route to dynamic, heterogeneous, and constraint-intensive counter-UAV task planning.
- Open Access
Long-Term QoT Forecasting in Dynamic Optical Networks via Decomposition-Driven Parallel Temporal Modeling
Accurate long-term forecasting of Quality of Transmission (QoT) is critical for the proactive operation and condition-aware management of dynamic elastic optical networks. However, the evolution of QoT is governed by multi-scale dynamics, including slow equipment aging, periodic operating variations, and short-term channel fluctuations, which a single temporal model struggles to capture jointly. To address this issue, we propose PA-TCN-Informer, a decomposition-driven parallel forecasting framework for long-horizon QoT prediction. The proposed framework first applies Seasonal-Trend decomposition using Loess (STL) to separate the Q-factor sequence into trend, seasonal, and residual components, and then employs Variational Mode Decomposition (VMD) to further resolve the residual into short-term fluctuation modes. The decomposed components, together with physical-layer monitoring features, are fed into a parallel TCN–Informer architecture, in which the TCN branch captures local temporal patterns while the Informer branch models long-range dependencies; the two streams are subsequently fused. We evaluate the proposed framework through Optuna-based hyperparameter optimization, STL/VMD sensitivity analysis, decomposition-method comparison, multi-seed baseline comparison with statistical testing, and zero-shot leave-one-dataset-out cross-domain evaluation. On the primary dataset, PA-TCN-Informer achieves the best overall forecasting accuracy among the compared models and reduces MAE by 2.2% relative to the serial TCN–Informer. In addition, the staged STL-VMD preprocessing alone yields a 60.8% reduction in MAE compared with raw inputs, confirming the value of physically interpretable multi-scale decomposition. In the zero-shot cross-domain setting, PA-TCN-Informer remains competitive across target domains. These results demonstrate that the proposed framework provides an effective and interpretable approach to QoT forecasting, and they further indicate that topology-aware modeling is a promising direction for improving cross-domain generalization.
Open AccessLong-Term QoT Forecasting in Dynamic Optical Networks via Decomposition-Driven Parallel Temporal Modeling
Accurate long-term forecasting of Quality of Transmission (QoT) is critical for the proactive operation and condition-aware management of dynamic elastic optical networks. However, the evolution of QoT is governed by multi-scale dynamics, including slow equipment aging, periodic operating variations, and short-term channel fluctuations, which a single temporal model struggles to capture jointly. To address this issue, we propose PA-TCN-Informer, a decomposition-driven parallel forecasting framework for long-horizon QoT prediction. The proposed framework first applies Seasonal-Trend decomposition using Loess (STL) to separate the Q-factor sequence into trend, seasonal, and residual components, and then employs Variational Mode Decomposition (VMD) to further resolve the residual into short-term fluctuation modes. The decomposed components, together with physical-layer monitoring features, are fed into a parallel TCN–Informer architecture, in which the TCN branch captures local temporal patterns while the Informer branch models long-range dependencies; the two streams are subsequently fused. We evaluate the proposed framework through Optuna-based hyperparameter optimization, STL/VMD sensitivity analysis, decomposition-method comparison, multi-seed baseline comparison with statistical testing, and zero-shot leave-one-dataset-out cross-domain evaluation. On the primary dataset, PA-TCN-Informer achieves the best overall forecasting accuracy among the compared models and reduces MAE by 2.2% relative to the serial TCN–Informer. In addition, the staged STL-VMD preprocessing alone yields a 60.8% reduction in MAE compared with raw inputs, confirming the value of physically interpretable multi-scale decomposition. In the zero-shot cross-domain setting, PA-TCN-Informer remains competitive across target domains. These results demonstrate that the proposed framework provides an effective and interpretable approach to QoT forecasting, and they further indicate that topology-aware modeling is a promising direction for improving cross-domain generalization.
- Top Cited PapersOpen Access
- 14 January 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (2) , 266
Two-Stage Fine-Tuning of Large Vision-Language Models with Hierarchical Prompting for Few-Shot Object Detection in Remote Sensing Images
Few-shot object detection (FSOD) in high-resolution remote sensing (RS) imagery remains challenging due to scarce annotations, large intra-class variability, and high visual similarity between categories, which together limit the generalization ability of convolutional neural network (CNN)-based detectors. To address this issue, we explore leveraging large vision-language models (LVLMs) for FSOD in RS. We propose a two-stage, parameter-efficient fine-tuning framework with hierarchical prompting that adapts Qwen3-VL for object detection. In the first stage, low-rank adaptation (LoRA) modules are inserted into the vision and text encoders and trained jointly with a Detection Transformer (DETR)-style detection head on fully annotated base classes under three-level hierarchical prompts. In the second stage, the vision LoRA parameters are frozen, the text encoder is updated using -shot novel-class samples, and the detection head is partially frozen, with selected components refined using the same three-level hierarchical prompting scheme. To preserve base-class performance and reduce class confusion, we further introduce knowledge distillation and semantic consistency losses. Experiments on the DIOR and NWPU VHR-10.v2 datasets show that the proposed method consistently improves novel-class performance while maintaining competitive base-class accuracy and surpasses existing baselines, demonstrating the effectiveness of integrating hierarchical semantic reasoning into LVLM-based FSOD for RS imagery.- 14 January 2026
- Published by MDPI AG in Remote Sensing
- Vol. 18 (2) , 266
Top Cited PapersOpen AccessTwo-Stage Fine-Tuning of Large Vision-Language Models with Hierarchical Prompting for Few-Shot Object Detection in Remote Sensing Images
Few-shot object detection (FSOD) in high-resolution remote sensing (RS) imagery remains challenging due to scarce annotations, large intra-class variability, and high visual similarity between categories, which together limit the generalization ability of convolutional neural network (CNN)-based detectors. To address this issue, we explore leveraging large vision-language models (LVLMs) for FSOD in RS. We propose a two-stage, parameter-efficient fine-tuning framework with hierarchical prompting that adapts Qwen3-VL for object detection. In the first stage, low-rank adaptation (LoRA) modules are inserted into the vision and text encoders and trained jointly with a Detection Transformer (DETR)-style detection head on fully annotated base classes under three-level hierarchical prompts. In the second stage, the vision LoRA parameters are frozen, the text encoder is updated using -shot novel-class samples, and the detection head is partially frozen, with selected components refined using the same three-level hierarchical prompting scheme. To preserve base-class performance and reduce class confusion, we further introduce knowledge distillation and semantic consistency losses. Experiments on the DIOR and NWPU VHR-10.v2 datasets show that the proposed method consistently improves novel-class performance while maintaining competitive base-class accuracy and surpasses existing baselines, demonstrating the effectiveness of integrating hierarchical semantic reasoning into LVLM-based FSOD for RS imagery. - Open Access
- 23 December 2025
- Published by MDPI AG in Preprints.org
Two-Stage Fine-Tuning of Large Vision-Language Models with Hierarchical Prompting for Few-Shot Object Detection in Remote Sensing Images
Few-shot object detection (FSOD) in high-resolution remote sensing (RS) imagery remains challenging due to scarce annotations, large intra-class variability, and high visual similarity between categories, which together limit the generalization ability of convolutional neural network (CNN)-based detectors. To address this issue, we explore leveraging large vision-language models (LVLMs) for FSOD in RS. We propose a two-stage, parameter-efficient fine-tuning framework with hierarchical prompting that adapts Qwen3-VL for object detection. In the first stage, low-rank adaptation (LoRA) modules are inserted into the vision and text encoders and trained jointly with a Detection Transformer (DETR)-style detection head on fully annotated base classes under three-level hierarchical prompts. In the second stage, the vision LoRA parameters are frozen, the text encoder is updated using K-shot novel-class samples, and the detection head is partially frozen, with selected components refined using the same three-level hierarchical prompting scheme. To preserve base-class performance and reduce class confusion, we further introduce knowledge distillation and semantic consistency losses. Experiments on the DIOR and NWPU VHR-10.v2 datasets show that the proposed method consistently improves novel-class performance while maintaining competitive base-class accuracy and surpasses existing baselines, demonstrating the effectiveness of integrating hierarchical semantic reasoning into LVLM-based FSOD for RS imagery.
- 23 December 2025
- Published by MDPI AG in Preprints.org
Open AccessTwo-Stage Fine-Tuning of Large Vision-Language Models with Hierarchical Prompting for Few-Shot Object Detection in Remote Sensing Images
Few-shot object detection (FSOD) in high-resolution remote sensing (RS) imagery remains challenging due to scarce annotations, large intra-class variability, and high visual similarity between categories, which together limit the generalization ability of convolutional neural network (CNN)-based detectors. To address this issue, we explore leveraging large vision-language models (LVLMs) for FSOD in RS. We propose a two-stage, parameter-efficient fine-tuning framework with hierarchical prompting that adapts Qwen3-VL for object detection. In the first stage, low-rank adaptation (LoRA) modules are inserted into the vision and text encoders and trained jointly with a Detection Transformer (DETR)-style detection head on fully annotated base classes under three-level hierarchical prompts. In the second stage, the vision LoRA parameters are frozen, the text encoder is updated using K-shot novel-class samples, and the detection head is partially frozen, with selected components refined using the same three-level hierarchical prompting scheme. To preserve base-class performance and reduce class confusion, we further introduce knowledge distillation and semantic consistency losses. Experiments on the DIOR and NWPU VHR-10.v2 datasets show that the proposed method consistently improves novel-class performance while maintaining competitive base-class accuracy and surpasses existing baselines, demonstrating the effectiveness of integrating hierarchical semantic reasoning into LVLM-based FSOD for RS imagery.
- Open Access
- 17 July 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (14) , 7961
Deep Reinforcement Learning-Based Deployment Method for Emergency Communication Network
Emergency communication networks play a crucial role in disaster relief operations. Current automated deployment strategies based on rule-driven or heuristic algorithms struggle to adapt to the dynamic and heterogeneous network environments in disaster scenarios, while manual command deployment is constrained by personnel expertise and response time requirements, leading to suboptimal trade-offs between deployment efficiency and reliability. To address these challenges, this study proposes a novel deep reinforcement learning framework with a fully convolutional value network architecture, which achieves breakthroughs in multi-dimensional spatial decision-making through end-to-end feature extraction. This design effectively mitigates the “curse of dimensionality” inherent in traditional reinforcement learning methods for topology planning. Experimental results demonstrate that the proposed method effectively accomplishes the planning tasks of emergency communication hub elements, significantly improving deployment efficiency while maintaining robustness in complex environments.
- 17 July 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (14) , 7961
Open AccessDeep Reinforcement Learning-Based Deployment Method for Emergency Communication Network
Emergency communication networks play a crucial role in disaster relief operations. Current automated deployment strategies based on rule-driven or heuristic algorithms struggle to adapt to the dynamic and heterogeneous network environments in disaster scenarios, while manual command deployment is constrained by personnel expertise and response time requirements, leading to suboptimal trade-offs between deployment efficiency and reliability. To address these challenges, this study proposes a novel deep reinforcement learning framework with a fully convolutional value network architecture, which achieves breakthroughs in multi-dimensional spatial decision-making through end-to-end feature extraction. This design effectively mitigates the “curse of dimensionality” inherent in traditional reinforcement learning methods for topology planning. Experimental results demonstrate that the proposed method effectively accomplishes the planning tasks of emergency communication hub elements, significantly improving deployment efficiency while maintaining robustness in complex environments.
- Open Access
- 15 June 2023
- Published by MDPI AG in Electronics
- Vol. 12 (12) , 2692
Entity Linking Method for Chinese Short Texts with Multiple Embedded Representations
Entity linking, a crucial task in the realm of natural language processing, aims to link entity mentions in a text to their corresponding entities in the knowledge base. While long documents provide abundant contextual information, facilitating feature extraction for entity identification and disambiguation, entity linking in Chinese short texts presents significant challenges. This study introduces an innovative approach to entity linking within Chinese short texts, combining multiple embedding representations. It integrates embedding representations from both entities and relations in the knowledge graph triples, as well as embedding representations from the descriptive text of entities and relations, to enhance the performance of entity linking. The method also incorporates external semantic supplements to strengthen the model’s feature learning capabilities. The Multi-Embedding Representation–Bidirectional Encoder Representation from Transformers–Bidirectional Gated Recurrent Unit (MER-BERT-BiGRU) neural network model is employed for embedding learning. The precision, recall, and F1 scores reached 89.73%, 92.18%, and 90.94% respectively, demonstrating the effectiveness of our approach.
- 15 June 2023
- Published by MDPI AG in Electronics
- Vol. 12 (12) , 2692
Open AccessEntity Linking Method for Chinese Short Texts with Multiple Embedded Representations
Entity linking, a crucial task in the realm of natural language processing, aims to link entity mentions in a text to their corresponding entities in the knowledge base. While long documents provide abundant contextual information, facilitating feature extraction for entity identification and disambiguation, entity linking in Chinese short texts presents significant challenges. This study introduces an innovative approach to entity linking within Chinese short texts, combining multiple embedding representations. It integrates embedding representations from both entities and relations in the knowledge graph triples, as well as embedding representations from the descriptive text of entities and relations, to enhance the performance of entity linking. The method also incorporates external semantic supplements to strengthen the model’s feature learning capabilities. The Multi-Embedding Representation–Bidirectional Encoder Representation from Transformers–Bidirectional Gated Recurrent Unit (MER-BERT-BiGRU) neural network model is employed for embedding learning. The precision, recall, and F1 scores reached 89.73%, 92.18%, and 90.94% respectively, demonstrating the effectiveness of our approach.