Jianquan Ouyang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jianquan Ouyang has a total of 77 co-authors. Here are the top co-authors by number of shared publications:
Huanrong Tang
Xiangtan University
13 shared publications- Huanrong Tang
School of Computer Science, Xiangtan University, Xiangtan, Hunan, China
5 shared publications Tianming Liu
The University of Georgia
5 shared publications- Chenyang Jiang2 shared publications
- Henry Mukalazi Serugunda
School of Computer Science, Xiangtan University, Xiangtan, Hunan, Uganda and Department of Information Technology, School of Computing and Informatics Technology, Makerere University, Kampala, Uganda
2 shared publications - Huanrong Tang
Xiangtan University
2 shared publications
Latest Publications
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3760-3769
Performance-Adaptive Instruction-Template Assignment for Clinical Reasoning in Large Language Models
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3760-3769
Open AccessPerformance-Adaptive Instruction-Template Assignment for Clinical Reasoning in Large Language Models
- 19 July 2026
- Published by Springer Nature
- p. 529-540
Risk-Sensitive Distributional Reinforcement Learning for Robust Stratospheric Balloon Station-Keeping
- 19 July 2026
- Published by Springer Nature
- p. 529-540
Risk-Sensitive Distributional Reinforcement Learning for Robust Stratospheric Balloon Station-Keeping
- Open Access
- 19 July 2026
- Published by Springer Nature
- p. 324-335
ThermoCausalNet: A Physics-Constrained Dynamic Convolutional Framework for Short-Term Temperature Forecasting
- 19 July 2026
- Published by Springer Nature
- p. 324-335
Open AccessThermoCausalNet: A Physics-Constrained Dynamic Convolutional Framework for Short-Term Temperature Forecasting
- 19 July 2026
- Published by Springer Nature
- p. 516-527
FedHAM: Heterogeneity-Aware Mixup and Knowledge Distillation for Federated Learning
- 19 July 2026
- Published by Springer Nature
- p. 516-527
FedHAM: Heterogeneity-Aware Mixup and Knowledge Distillation for Federated Learning
- 17 July 2026
- Published by Springer Nature
- p. 3-13
Triplet-Aware Sparse Interaction for Aspect Sentiment Triplet Extraction
- 17 July 2026
- Published by Springer Nature
- p. 3-13
Triplet-Aware Sparse Interaction for Aspect Sentiment Triplet Extraction
- 16 July 2026
- Published by Springer Nature
- p. 15-27
Adaptive Post-hoc Knowledge Consistency for Meteorological Prediction Under Extreme Uncertainty
- 16 July 2026
- Published by Springer Nature
- p. 15-27
Adaptive Post-hoc Knowledge Consistency for Meteorological Prediction Under Extreme Uncertainty
- 13 July 2026
- Published by Springer Nature
- p. 35-45
CryoHIRI: A Knowledge-Driven Framework for High-Resolution Structural Knowledge Discovery from Cryo-EM Images
- 13 July 2026
- Published by Springer Nature
- p. 35-45
CryoHIRI: A Knowledge-Driven Framework for High-Resolution Structural Knowledge Discovery from Cryo-EM Images
- 12 July 2026
- Published by Springer Nature
- p. 669-679
Knowledge-Driven 3D Reconstruction: Fusing Explicit Structural Priors with Latent Diffusion Models
- 12 July 2026
- Published by Springer Nature
- p. 669-679
Knowledge-Driven 3D Reconstruction: Fusing Explicit Structural Priors with Latent Diffusion Models
- 12 July 2026
- Published by Springer Nature
- p. 140-149
Difference-Aware Boundary Refinement for Named Entity Recognition
- 12 July 2026
- Published by Springer Nature
- p. 140-149
Difference-Aware Boundary Refinement for Named Entity Recognition
- Open Access
- 23 June 2026
- Published by Elsevier in SSRN Electronic Journal
A Lightweight Hybrid Network with Reinforcement Learning Optimization for Pig Detection and Counting
To address the challenge of balancing model lightweight design with high accuracy in automated pig counting for large-scale farming, this paper proposes a lightweight counting method based on a hybrid network architecture optimized with reinforcement learning (DQN). Using YOLO11n as the detection backbone, it incorporates Swin Transformer to enhance global perception and embeds CBAM to focus on key regions. The core innovation formulates learning rate adjustment as a sequential decision problem, adopting DQN for adaptive optimization, thereby improving convergence and performance under complex disturbances. Experiments show that while remaining lightweight, the model improves mAP50 and mAP50-95 by 0.8% and 1.3% over the baseline, demonstrating its accuracy and edge deployment feasibility in complex farming environments.
- 23 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessA Lightweight Hybrid Network with Reinforcement Learning Optimization for Pig Detection and Counting
To address the challenge of balancing model lightweight design with high accuracy in automated pig counting for large-scale farming, this paper proposes a lightweight counting method based on a hybrid network architecture optimized with reinforcement learning (DQN). Using YOLO11n as the detection backbone, it incorporates Swin Transformer to enhance global perception and embeds CBAM to focus on key regions. The core innovation formulates learning rate adjustment as a sequential decision problem, adopting DQN for adaptive optimization, thereby improving convergence and performance under complex disturbances. Experiments show that while remaining lightweight, the model improves mAP50 and mAP50-95 by 0.8% and 1.3% over the baseline, demonstrating its accuracy and edge deployment feasibility in complex farming environments.