Robust semi-supervised classification based on data augmented online ELMs with deep features

Hu, Xiaochang; Zeng, Yujun; Xu, Xin; Zhou, Sihang; Liu, Li

doi:10.1016/j.knosys.2021.107307

Cited by 11 publications

(2 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consistency regularization assumes that the model's predictions should remain consistent when small perturbations are introduced to the unlabeled data. Commonly used perturbations include image augmentations [27][28][29], adversarial training [30,31], and model-level perturbations [11,32]. On the other hand, self-training methods treat the predictions on unlabeled data as pseudo labels, which are then incorporated into the re-training process.…”

Section: B Semi-supervised Learning In Classificationmentioning

confidence: 99%

Weighted Pseudo-Labels and Bounding Boxes for Semisupervised SAR Target Detection

Tian,

Wang,

Zhou

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Synthetic aperture radar (SAR) image target detection methods based on semi-supervised learning, such as the mean teacher framework, have shown promise in diminishing the issue of limited labeled data. However, several challenges exist in current methods. Firstly, data augmentation techniques designed for optical images may not be suitable for SAR images due to differences in imaging methods. Additionally, the contribution of pseudo labels remains constant during the initial retraining stage can lead to degradation in prediction results. Moreover, the low quality of predicted bounding boxes poses a challenge for effective retraining. To address these challenges, we propose an end-toend semi-supervised detection method based on the mean teacher framework. To enhance the robustness of training, we firstly introduce SAR-specific data augmentation techniques, including multiplicative noise, which effectively increase the diversity of training samples. Secondly, we propose a method that weights the losses of pseudo-labeled data using a hard-sigmoid function, gradually improving the importance of pseudo-labeled data during retraining, thereby alleviating their potential negative impact on the training process. Finally, we propose an IoUaware subnetwork to incorporate high-quality pseudo-labeled bounding boxes into retraining, allowing them to contribute to network adjustments while mitigating the impact of low-quality samples. Experimental evaluations on publicly available SAR image datasets demonstrate the effectiveness of our proposed method in improving the target detection capability of semisupervised SAR target detection.

show abstract

Section: B Semi-supervised Learning In Classificationmentioning

confidence: 99%

Weighted Pseudo-Labels and Bounding Boxes for Semisupervised SAR Target Detection

Tian,

Wang,

Zhou

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…ELM, as a single hidden layer feedforward neural network algorithm, was widely used in image classification [13,14], data label classification [15], fingerprint classification [16] and other fields, with good learning efficiency and generalization performance. Xiao et al [17] classified 180 samples of 6 kinds of construction waste obtained by hyperspectral technology, and the accuracy of ELM can reach 100%, showing a strong classification ability.…”

Section: Introductionmentioning

confidence: 99%

Research on Rosewood Micro Image Classification Method Based on Feature Fusion and ELM

Yang¹,

Gao²,

Zhang³

et al. 2022

Journal of Renewable Materials

View full text Add to dashboard Cite

Rosewood is a kind of high-quality and precious wood in China. The correct identification of rosewood species is of great significance to the import and export trade and species identification of furniture materials. In this paper, micro CT was used to obtain the micro images of cross sections, radial sections and tangential sections of 24 kinds of rosewood, and the data sets were constructed. PCA method was used to reduce the dimension of four features including logical binary pattern, local configuration pattern, rotation invariant LBP, uniform LBP. These four features and one feature not reducing dimension (rotation invariant uniform LBP) was fused with Gray Level Co-Occurrence Matrix and Tamura features, respectively, a total of five fused features LBP+GLCM+Tamura, LCP +GLCM+Tamura, LBP u2 P;R +GLCM+Tamura, LBP ri P;R +GLCM+Tamura and LBP riu2 P;R +GLCM+Tamura were obtained. The five fused features were classified by extreme learning machine and BP neural network. The classification effect of feature LBP u2 P;R +GLCM+Tamura combined with extreme learning machine was the best, and the classification accuracy of cross, radial and tangential sections reached 100%, 97.63% and 94.72%, respectively, which is 0.83%, 2.77% and 5.70% higher than that of BP neural network. The classification running time of ELM is less than 1 s, and the classification efficiency is high. In conclusion, the LBP u2 P;R +GLCM+Tamura method combined with extreme learning machine can be used as a quick and accurate classifier, providing an efficient and feasible classification method of rosewood.

show abstract

A Semi-supervised Classification Method for 6G Remote Sensing Images Based on Pseudo-label and False Representation Recognition

Meng

Liang

Liu

2023

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

Robust semi-supervised classification based on data augmented online ELMs with deep features

Cited by 11 publications

References 47 publications

Weighted Pseudo-Labels and Bounding Boxes for Semisupervised SAR Target Detection

Weighted Pseudo-Labels and Bounding Boxes for Semisupervised SAR Target Detection

Research on Rosewood Micro Image Classification Method Based on Feature Fusion and ELM

A Semi-supervised Classification Method for 6G Remote Sensing Images Based on Pseudo-label and False Representation Recognition

Contact Info

Product

Resources

About