DCBT-Net: Training Deep Convolutional Neural Networks With Extremely Noisy Labels

Abstract

Obtaining data with correct labels is crucial to attain the state-of-the-art performance of Convolutional Neural Network (CNN) models. However, labeling datasets is significantly time-consuming and expensive process because it requires expert knowledge in a particular domain. Therefore, real-life datasets often exhibit incorrect labels due to the involvement of nonexperts in the data-labeling process. Consequently, there are many cases of incorrectly labeled data in the wild. Although the issue of poorly labeled datasets has been studied, the existing methods are complex and difficult to reproduce. Thus, in this study, we proposed a simpler algorithm called "Deep Clean Before Training Net" (DCBT-Net) that is based on cleaning wrongly labeled data points using the information from eigenvalues of the Laplacian matrix obtained from similarities between the data samples. The cleaned data were trained using deep CNN (DCNN) to attain the state-of-the-art results. This system achieved better performance than the existing approaches. In conducted experiments, the performance of the DCBT-Net was tested on three commercially available datasets, namely, Modified National Institute of Standards a nd Technology (MNIST) database of handwritten digits, Canadian Institute for Advanced Research (CIFAR) and WebVision1000 datasets. The proposed method achieved better results when assessed using several evaluation metrics compared with the existing state-of-the-art methods. Specifically, the DCBT-Net attained an average 15%, 20%, and 3% increase in accuracy score using MNIST database, CIFAR-10 dataset, and WebVision dataset, respectively. Also, the proposed approach demonstrated better results in specificit y, sensitivity, positive predictive value, and negative predictive value evaluation metrics. INDEX TERMS Clustering, deep convolutional neural networks, eigenvalues and eigenvectors, image classification, noisy (corrupted) labels.

A Deep Learning Model for Bacterial Classification Using Big Transfer (BiT)

“…In this research, we apply the Graph Laplacian-based sample weighting method to prevent noisy labels. First, inspired by Olimov et al [17], we applied the Graph Laplacian method to find potential noisy labels for each class of inputs. However, unlike [17], instead of deleting all noisy data from the training sets, we assigned different sample weights to those detected noisy data.…”

“…First, inspired by Olimov et al [17], we applied the Graph Laplacian method to find potential noisy labels for each class of inputs. However, unlike [17], instead of deleting all noisy data from the training sets, we assigned different sample weights to those detected noisy data. (ii) Similarity Matrix: A similarity matrix (𝑆) is a square matrix that describes pairwise similarity or dissimilarity between two nodes.…”

A customized deep learning framework for skin lesion classification using dermoscopic images

“…Deep learning models have shown competitive performance compared to dermatologists demonstrated in Reference 14. In this regard, CNN 15,16 quickly becomes a choice in examining dermoscopic images. [17][18][19][20][21][22][23][24][25] For better classification performance, accurate lesion area extraction is very important.…”

Consecutive multiscale feature learning-based image classification model

“…Image classification involves the extraction of useful features from a digital image and the classification of the image into one of the pre-defined classes based on the extracted features 12 , 13 . Manual verification and classification of digital images can be a laborious and monotonous process; thus, automating the image analysis process by using image classification methods is more efficient and less time-consuming 14 , 15 . Recent advances in these methods have facilitated the usage of image classification in several real-world applications, such as medical imaging 16 , 17 , face recognition 18 , human activity recognition 19 , and traffic control systems 20 , 21 .…”