Active Learning Plus Deep Learning Can Establish Cost-Effective and Robust Model for Multichannel Image: A Case on Hyperspectral Image Classification

Abstract

Relying on large scale labeled datasets, deep learning has achieved good performance in image classification tasks. In agricultural and biological engineering, image annotation is time-consuming and expensive. It also requires annotators to have technical skills in specific areas. Obtaining the ground truth is difficult because natural images are expensive. In addition, images in these areas are usually stored as multichannel images, such as computed tomography (CT) images, magnetic resonance images (MRI), and hyperspectral images (HSI). In this paper, we present a framework using active learning and deep learning for multichannel image classification. We use three active learning algorithms, including least confidence, margin sampling, and entropy, as the selection criteria. Based on this framework, we further introduce an “image pool” to make full advantage of images generated by data augmentation. To prove the availability of the proposed framework, we present a case study on agricultural hyperspectral image classification. The results show that the proposed framework achieves better performance compared with the deep learning model. Manual annotation of all the training sets achieves an encouraging accuracy. In comparison, using active learning algorithm of entropy and image pool achieves a similar accuracy with only part of the whole training set manually annotated. In practical application, the proposed framework can remarkably reduce labeling effort during the model development and upadting processes, and can be applied to multichannel image classification in agricultural and biological engineering.

Computed Tomography Image Features under Deep Learning Algorithm Applied in Staging Diagnosis of Bladder Cancer and Detection on Ceramide Glycosylation

“…In this study, the deep learning algorithms were applied to CT images to detect bladder cancer staging, the convolutional neural network (CNN) algorithm was adopted to extract features of tumor regions or bladder wall regions, and the models were trained to quickly and accurately classify or segment bladder tumors. Among them, the You Only Look Once (YOLO) target detection algorithm based on deep learning used the CNN model to extract the features of the predicted region and perform classification and recognition [8,9].…”

NDAMA: A Novel Deep Autoencoder and Multivariate Analysis Approach for IoT-Based Methane Gas Leakage Detection

“…Next, in Fig 6 (b), we show a relationship between the number of trees in the model and the depth of each tree. We created a grid of 9 different n estimators' values (100 to 500) and 6 different max depth values (2,4,6,8,10,12), and each combination was evaluated using 10-fold cross-validation. A total of 9 × 6 × 10 or 540 models were trained and evaluated.…”

“…The proposed method consists of three main modules: data preprocessing, data labeling, and predictive analysis. The data preprocessing module removes outliers using the deep autoencoder (DAE) reconstruction error (RE) [6] and normalizes the data using OrdinalEncoder (OE) transformation techniques. The data labeling module selects only natural gas (NG) CH4 [7,8] data from the preprocessed data, divides it into groups using the k-means clustering algorithm, and classifies the data according to that group.…”

Discrete Wavelet Analysis: A Mighty Approach for Image Segmentation

“…Numerous algorithms and techniques have been developed for image segmentation, ranging from traditional methods like thresholding, edge detection [12], and region growing [13] to more advanced approaches such as clustering [14],…”