학술논문
An Architecture Combining Convolutional Neural Network (CNN) with Batch Normalization for Apparel Image Classification
Document Type
Conference
Author
Source
2020 IEEE International Symposium on Sustainable Energy, Signal Processing and Cyber Security (iSSSC) Sustainable Energy, Signal Processing and Cyber Security (iSSSC), 2020 IEEE International Symposium on. :1-6 Dec, 2020
Subject
Language
Abstract
In recent past, Convolutional Neural Networks (CNN) have been utilized in kind of areas, including style order. Web-based media, web-based business, and legitimate code are widely appropriate during this field. CNNs are proficient to prepare and situated to offer the preeminent exact prompts tackling world issues. In this paper, CNN based profound learning a cutting-edge (state-of-the-Art) model proposed for classification of Apparel images of Fashion MNIST dataset. Different CNN models proposed utilizing Dropout and Batch Normalization (BN) with Early Stopping to quicken learning measure and forestall overfitting. In view of correlations it is seen that proposed models improved accuracy and precision over the common best in class frameworks given in literature.