Haiyang Yang
Research Subjects
Haiyang Yang has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
Hongshuai Lei
5 shared publicationsDaining Fang
3 shared publicationsXinyu Li
Institute of Advanced Structure Technology, Beijing Institute of Technology, Beijing, 100081, PR China
3 shared publications- Zhong Zhang3 shared publications
Changxian Wang
2 shared publications- Chongjun Wang
Nanjing University
2 shared publications
Latest Publications
- Top Cited Papers
- 27 January 2025
- Published by Elsevier in Reliability Engineering & System Safety
- Vol. 257, 110847
Novel domain-adaptive Wasserstein generative adversarial networks for early bearing fault diagnosis under various conditions
- 27 January 2025
- Published by Elsevier in Reliability Engineering & System Safety
- Vol. 257, 110847
Top Cited PapersNovel domain-adaptive Wasserstein generative adversarial networks for early bearing fault diagnosis under various conditions
- 20 December 2023
- Published by Springer Nature in Machine Learning
- Vol. 113 (6) , 3611-3631
Generation, augmentation, and alignment: a pseudo-source domain based method for source-free domain adaptation
- 20 December 2023
- Published by Springer Nature in Machine Learning
- Vol. 113 (6) , 3611-3631
Generation, augmentation, and alignment: a pseudo-source domain based method for source-free domain adaptation
- Open Access
Crashworthy optimization of skeleton-filled FRP tubes based on back propagation neural network
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 13, 1-6
Speed Control of Permanent Magnet Synchronous Motor Based on Improved Active Disturbance Rejection Control
This study examines an improved Active Disturbance Rejection Controller (I-ADRC) to enhance the Surface Mounted Permanent Magnet Synchronous Motor’s (SPMSM) capacity to regulate speed. A nonlinear function with increased smoothness is utilized to develop first-order active disturbance rejection controllers for the loops of speed and current, respectively, followed by the creation of a load torque observer and its integration into the speed controller’s disturbance compensation. The emulation findings demonstrate that compared to the PI controller and the conventional ADRC controller, the PMSM speed control system based on the enhanced ADRC has reduced current fluctuation, smaller overshoot, greater anti-disturbance capability, and greater robustness.
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 13, 1-6
Speed Control of Permanent Magnet Synchronous Motor Based on Improved Active Disturbance Rejection Control
This study examines an improved Active Disturbance Rejection Controller (I-ADRC) to enhance the Surface Mounted Permanent Magnet Synchronous Motor’s (SPMSM) capacity to regulate speed. A nonlinear function with increased smoothness is utilized to develop first-order active disturbance rejection controllers for the loops of speed and current, respectively, followed by the creation of a load torque observer and its integration into the speed controller’s disturbance compensation. The emulation findings demonstrate that compared to the PI controller and the conventional ADRC controller, the PMSM speed control system based on the enhanced ADRC has reduced current fluctuation, smaller overshoot, greater anti-disturbance capability, and greater robustness.
- 17 March 2023
- Published by Springer Nature
- p. 642-658
InCo: Intermediate Prototype Contrast for Unsupervised Domain Adaptation
- 17 March 2023
- Published by Springer Nature
- p. 642-658
InCo: Intermediate Prototype Contrast for Unsupervised Domain Adaptation
- Open Access
Saliency Guided Contrastive Learning on Scene Images
Self-supervised learning holds promise in leveraging large numbers of unlabeled data. However, its success heavily relies on the highly-curated dataset, e.g., ImageNet, which still needs human cleaning. Directly learning representations from less-curated scene images is essential for pushing self-supervised learning to a higher level. Different from curated images which include simple and clear semantic information, scene images are more complex and mosaic because they often include complex scenes and multiple objects. Despite being feasible, recent works largely overlooked discovering the most discriminative regions for contrastive learning to object representations in scene images. In this work, we leverage the saliency map derived from the model's output during learning to highlight these discriminative regions and guide the whole contrastive learning. Specifically, the saliency map first guides the method to crop its discriminative regions as positive pairs and then reweighs the contrastive losses among different crops by its saliency scores. Our method significantly improves the performance of self-supervised learning on scene images by +1.1, +4.3, +2.2 Top1 accuracy in ImageNet linear evaluation, Semi-supervised learning with 1% and 10% ImageNet labels, respectively. We hope our insights on saliency maps can motivate future research on more general-purpose unsupervised representation learning from scene data.
Open AccessSaliency Guided Contrastive Learning on Scene Images
Self-supervised learning holds promise in leveraging large numbers of unlabeled data. However, its success heavily relies on the highly-curated dataset, e.g., ImageNet, which still needs human cleaning. Directly learning representations from less-curated scene images is essential for pushing self-supervised learning to a higher level. Different from curated images which include simple and clear semantic information, scene images are more complex and mosaic because they often include complex scenes and multiple objects. Despite being feasible, recent works largely overlooked discovering the most discriminative regions for contrastive learning to object representations in scene images. In this work, we leverage the saliency map derived from the model's output during learning to highlight these discriminative regions and guide the whole contrastive learning. Specifically, the saliency map first guides the method to crop its discriminative regions as positive pairs and then reweighs the contrastive losses among different crops by its saliency scores. Our method significantly improves the performance of self-supervised learning on scene images by +1.1, +4.3, +2.2 Top1 accuracy in ImageNet linear evaluation, Semi-supervised learning with 1% and 10% ImageNet labels, respectively. We hope our insights on saliency maps can motivate future research on more general-purpose unsupervised representation learning from scene data.
- 23 October 2022
- Published by Springer Nature
- p. 151-168
Domain Invariant Masked Autoencoders for Self-supervised Learning from Multi-domains
- 23 October 2022
- Published by Springer Nature
- p. 151-168
Domain Invariant Masked Autoencoders for Self-supervised Learning from Multi-domains
- 14 October 2022
- Published by Elsevier in Composite Structures
- Vol. 303
Failure mechanisms of fluted-core sandwich composite panels under uniaxial compression
- 14 October 2022
- Published by Elsevier in Composite Structures
- Vol. 303
Failure mechanisms of fluted-core sandwich composite panels under uniaxial compression
- 16 June 2022
- Published by Elsevier in Composites Science and Technology
- Vol. 227
Radar-stealth and load-bearing corrugated sandwich structures with superior environmental adaptability
- 16 June 2022
- Published by Elsevier in Composites Science and Technology
- Vol. 227
Radar-stealth and load-bearing corrugated sandwich structures with superior environmental adaptability
- 17 February 2022
- Published by Elsevier in Composites Science and Technology
- Vol. 221
Load-bearing capacity and failure mechanism of integrated fluted-core composite sandwich cylinders
- 17 February 2022
- Published by Elsevier in Composites Science and Technology
- Vol. 221
Load-bearing capacity and failure mechanism of integrated fluted-core composite sandwich cylinders