Analysis of Residual Block in the ResNet for Image Classification

Xiayuan Jin

2023

Abstract

Image recognition is of paramount importance in our contemporary world, with diverse applications across domains such as traffic control, medical diagnosis, educational tools, and workplace automation. Its impact is profound and multifaceted. This study highlights Resnet’s effectiveness in building robust deep-learning models for image classification. Through the integration of residual blocks, residual network (ResNet) overcomes challenges like vanishing gradients, enabling the training of very deep networks. Experiments on the CIFAR-10 dataset showcase ResNet’s impressive accuracy in image recognition, with loss fluctuations mitigated via hyperparameter tuning. ResNet excels in feature extraction and precise image classification. The important topic of this research is trying to figure out the efficiency comparing convolutional neural network (CNN) and Resnet using Resnet’s residual block to find out the difference of parameter changes the accuracy of the model, by inducting Resnet, the performance of the model behaves much better, and solve the problem of gradient vanishment, Resnet plays a pivotal role in image classification by enabling the training of very deep neural networks, enhancing feature extraction, and achieving state-of-the-art accuracy in various visual recognition tasks.

Paper Citation

in Harvard Style

Jin X. (2023). Analysis of Residual Block in the ResNet for Image Classification. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 246-250. DOI: 10.5220/0012800400003885

in Bibtex Style

@conference{daml23,
author={Xiayuan Jin},
title={Analysis of Residual Block in the ResNet for Image Classification},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={246-250},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012800400003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Analysis of Residual Block in the ResNet for Image Classification
SN - 978-989-758-705-4
AU - Jin X.
PY - 2023
SP - 246
EP - 250
DO - 10.5220/0012800400003885
PB - SciTePress