A Robust Hierarchical Neural Network Methodology for Improved Image Classification Performance

Along with the strong representation of the convolutional neural network (CNN), image classification tasks have achieved considerable progress. However, majority of works focus on designing complicated and redundant architectures for extracting informative features to improve classification performance. In this study, we concentrate on rectifying the incomplete outputs of CNN. To be concrete, we propose an innovative image classification method based on Label Rectification Learning (LRL) through kernel extreme learning machine (KELM). It mainly consists of two steps: (1) preclassification, extracting incomplete labels through a pretrained CNN, and (2) label rectification, rectifying the generated incomplete labels by the KELM to obtain the rectified labels. Experiments conducted on publicly available datasets demonstrate the effectiveness of our method. Notably, our method is extensible which can be easily integrated with off-the-shelf networks for improving performance.

Download Full-text

AdaBoostNet: An Efficient Hierarchical Neural Network for Image Classification

2019 IEEE International Conference on Signal, Information and Data Processing (ICSIDP) ◽

10.1109/icsidp47821.2019.9173507 ◽

2019 ◽

Author(s):

Shichao Zhou ◽

Baojun Zhao ◽

Linbo Tang ◽

Donglin Jing ◽

Yu Pan ◽

...

Keyword(s):

Neural Network ◽

Image Classification ◽

Hierarchical Neural Network

Download Full-text

Fuzzy Superpixels Based Semi-Supervised Similarity-Constrained CNN for PolSAR Image Classification

Remote Sensing ◽

10.3390/rs12101694 ◽

2020 ◽

Vol 12 (10) ◽

pp. 1694 ◽

Cited By ~ 1

Author(s):

Yuwei Guo ◽

Zhuangzhuang Sun ◽

Rong Qu ◽

Licheng Jiao ◽

Fang Liu ◽

...

Keyword(s):

Neural Network ◽

Deep Learning ◽

Computational Complexity ◽

Convolutional Neural Network ◽

Image Classification ◽

Classification Performance ◽

Unlabeled Data ◽

Training Set ◽

Supervised Methods ◽

High Computational Complexity

Recently, deep learning has been highly successful in image classification. Labeling the PolSAR data, however, is time-consuming and laborious and in response semi-supervised deep learning has been increasingly investigated in PolSAR image classification. Semi-supervised deep learning methods for PolSAR image classification can be broadly divided into two categories, namely pixels-based methods and superpixels-based methods. Pixels-based semi-supervised methods are liable to be affected by speckle noises and have a relatively high computational complexity. Superpixels-based methods focus on the superpixels and ignore tiny detail-preserving represented by pixels. In this paper, a Fuzzy superpixels based Semi-supervised Similarity-constrained CNN (FS-SCNN) is proposed. To reduce the effect of speckle noises and preserve the details, FS-SCNN uses a fuzzy superpixels algorithm to segment an image into two parts, superpixels and undetermined pixels. Moreover, the fuzzy superpixels algorithm can also reduce the number of mixed superpixels and improve classification performance. To exploit unlabeled data effectively, we also propose a Similarity-constrained Convolutional Neural Network (SCNN) model to assign pseudo labels to unlabeled data. The final training set consists of the initial labeled data and these pseudo labeled data. Three PolSAR images are used to demonstrate the excellent classification performance of the FS-SCNN method with data of limited labels.

Download Full-text

Rider and Sunflower optimization-driven neural network for image classification

Web Intelligence ◽

10.3233/web-210454 ◽

2021 ◽

Vol 19 (1-2) ◽

pp. 41-61

Author(s):

Hanumantha Rao Nadendla ◽

A. Srikrishna ◽

K. Gangadhara Rao

Keyword(s):

Neural Network ◽

Machine Learning ◽

Image Classification ◽

Optimization Algorithm ◽

Pattern Analysis ◽

Classification Performance ◽

Optimization Approach ◽

Shape Decomposition ◽

Maximal Sensitivity ◽

Sensitivity Specificity

Image classification is the classical issue in computer vision, machine learning, and image processing. The image classification is measured by differentiating the image into the prescribed category based on the content of the vision. In this paper, a novel classifier named RideSFO-NN is developed for image classification. The proposed method performs the image classification by undergoing two steps, namely feature extraction and classification. Initially, the images from various sources are provided to the proposed Weighted Shape-Size Pattern Spectra for pattern analysis. From the pattern analysis, the significant features are obtained for the classification. Here, the proposed Weighted Shape-Size Pattern Spectra is designed by modifying the gray-scale decomposition with Weight-Shape decomposition. Then, the classification is done based on Neural Network (NN) classifier, which is trained using an optimization approach. The optimization will be done by the proposed Ride Sunflower optimization (RideSFO) algorithm, which is the integration of Rider optimization algorithm (ROA), and Sunflower optimization algorithm (SFO). Finally, the image classification performance is evaluated using RideSFO-NN based on sensitivity, specificity, and accuracy. The developed RideSFO-NN method achieves the maximal accuracy of 94%, maximal sensitivity of 93.87%, and maximal specificity of 90.52% based on K-Fold.

Download Full-text

Classification of Similar Sports Images Using Convolutional Neural Network with Hyper-Parameter Optimization

Applied Sciences ◽

10.3390/app10238494 ◽

2020 ◽

Vol 10 (23) ◽

pp. 8494

Author(s):

Vili Podgorelec ◽

Špela Pečnik ◽

Grega Vrbančič

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

Image Classification ◽

Parameter Optimization ◽

Classification Performance ◽

Training Data ◽

Fine Tuning ◽

American Football ◽

Massive Scale

With the exponential growth of the presence of sport in the media, the need for effective classification of sports images has become crucial. The traditional approaches require carefully hand-crafted features, which make them impractical for massive-scale data and less accurate in distinguishing images that are very similar in appearance. As the deep learning methods can automatically extract deep representation of training data and have achieved impressive performance in image classification, our goal was to apply them to automatic classification of very similar sports disciplines. For this purpose, we developed a CNN-TL-DE method for image classification using the fine-tuning of transfer learning for training a convolutional neural network model with the use of hyper-parameter optimization based on differential evolution. Through the automatic optimization of neural network topology and essential training parameters, we significantly improved the classification performance evaluated on a dataset composed from images of four similar sports—American football, rugby, soccer, and field hockey. The analysis of interpretable representation of the trained model additionally revealed interesting insights into how our model perceives images which contributed to a greater confidence in the model prediction. The performed experiments showed our proposed method to be a very competitive image classification method for distinguishing very similar sports and sport situations.

Download Full-text