Stochastic configuration networks for imbalanced data classification

Dai, Wei; Ning, Chuanfeng; Nan, Jing; Wang, Dianhui

doi:10.1007/s13042-022-01565-z

Cited by 10 publications

(3 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep binary descriptors (DeepBit) [19] uses VGGNet [20] to extract the features of images and learns the hashing codes with a combined object function of quantization loss, balanced regularization and rotation invariant objective. Stochastic generative hashing (SGH) [21] learns hashing codes by minimum description length principle so as to maximally compress the dataset as well as regenerate outputs from the codes. Semantic structure-based unsupervised hashing (SSDH) [22] uses two half Gaussian distributions to estimate pairwise cosine distances of data points and assign any two data points with obviously smaller distance as semantically similar pair.…”

Section: A Hashing Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing Non-Differentiable Metrics for Hashing

et al. 2021

View full text Add to dashboard Cite

Image hashing embeds the image to binary codes which can boost the efficiency of approximately nearest neighbors search. F-measure is a widely-used metric for evaluating the performance of hashing methods. However, it is non-differentiable and hence it has not been used as an object function for hashing. Heuristic algorithms, e.g. evolutionary computation and particle swarm optimization (PSO), are good at optimizing non-differentiable objectives, while they are inefficient in very high-dimensional variables which are commonly used in hashing models. To address this contradict, we propose a scheme to bridge hashing methods and F-measure objective using PSO. The hashing methods are used to generate real-valued codes for images and then the parameters of quantization procedure are optimized by PSO. Our scheme can incorporate a wide range of hashing methods, heuristic optimization algorithms and nondifferentiable metrics. Experimental results demonstrate that our scheme can be used to further improve the performance of existing hashing methods.

show abstract

Section: A Hashing Methodsmentioning

confidence: 99%

“…We incorporate five state-of-the-art traditional hashing methods to our scheme, i.e. ITQ [3], GHS [34], IMH [12], [21] and SSDH [22]. ITQ, GHS and IMH are shallow hashing methods, while SGH and SSDH adopts deep neural networks.…”

Section: B Baselinesmentioning

confidence: 99%

Optimizing Non-Differentiable Metrics for Hashing

et al. 2021

View full text Add to dashboard Cite

show abstract

“…In order to extract more information through DNNs, some researchers have introduced trendy self-supervised techniques in deep unsupervised hashing, such as Autoencoders in SGH [9], TBH [10] and BDNN [11], and the application of Generative Adversarial Networks (GAN) in HashGAN [12], BGAN [13] and BinGAN [14] etc.…”

Section: Deep Self-supervised Image Retrievalmentioning

confidence: 99%

Multi-view self-supervised product quantization for robust fabric image retrieval

Zhuo,

Zhong,

Chen

2024

Fifteenth International Conference on Graphics and Image Processing (ICGIP 2023)

View full text Add to dashboard Cite

Currently, fabric image retrieval faces challenges such as the high cost of image annotation and its vulnerability to adversarial perturbations. To minimize manual supervision and enhance the robustness of the retrieval system, this study proposes a robust deep image retrieval algorithm using multi-view self-supervised product quantization for artificially generated fabric images. The method introduces a multi-view module, which includes two views enhanced by AutoAugment, an adversarial view and a high-frequency view of the unlabeled images. AutoAugment can generate more varied data variations, which allows the model to learn more about the different features and structures of the fabric texture; fabric images are usually of high complexity and diversity, and adding the adversarial sample into the model training can add more noise and variations, which is one of the best existing ways to defend against adversarial attacks; the high-frequency component can make the edges, details, and contrasts in the fabric image clearer. A robust cross quantized contrastive loss function is also designed to jointly learn codewords and deep visual descriptors by comparing multiple views, effectively increasing the model's robustness and generalization. The method's effectiveness is demonstrated by experimental results on multiple datasets, which can significantly improve the robustness of the retrieval system compared to other state-of-the-art retrieval algorithms. Our method presents a new approach for fabric image retrieval and has great significance for improving its performance.

show abstract

Activity recognition in rehabilitation training based on ensemble stochastic configuration networks

Jiao

Wang

et al. 2023

Neural Comput & Applic

View full text Add to dashboard Cite

Stochastic configuration networks for imbalanced data classification

Cited by 10 publications

References 41 publications

Optimizing Non-Differentiable Metrics for Hashing

Optimizing Non-Differentiable Metrics for Hashing

Multi-view self-supervised product quantization for robust fabric image retrieval

Activity recognition in rehabilitation training based on ensemble stochastic configuration networks

Contact Info

Product

Resources

About