Addressing misclassification in deep learning: A Merged Net approach
Software Impacts · 2023-09 · 17 · 100525
DOI: 10.1016/j.simpa.2023.100525 ↗Kurz gefasst
Fachartikel in der Zeitschrift Software Impacts, 2023-09. Verfasst von Vishal Meshram und 3 weiteren Autor:innen. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 16 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The purpose of this research is to propose a novel framework called "MNet: Merged Net" for fast and accurate fruit classification, which addresses the misclassification problem overlooked by many researchers. To achieve this, a dataset consisting of 12,000 color images of top fruits in India with "Good" and "Bad" quality labels was created and published. The dataset consists of a total of 12 classes, and two models are created based on the popular pretrained model InceptionV3. The principal results of the study show that the proposed framework achieved 99.92% accuracy and moderated the image misclassification problem. In addition, the experimental results demonstrate that the MFC_InceptionV3 model outperforms the InceptionV3 and FC_InceptionV3 models in terms of accuracy and misclassification problem. The major conclusion of this study is that the proposed framework, MNet, is a promising approach for fast and accurate fruit classification, which improves the accuracy and addresses the misclassification problem. The proposed dataset and framework can be used as a benchmark for future research in this field and can contribute to the farming business by providing a reliable and efficient technique for fruit classification.
Zitieren
APA
Meshram, V., Suryawanshi, Y., Meshram, V., & Patil, K. (2023). Addressing misclassification in deep learning: A Merged Net approach. Software Impacts, 17, 100525. https://doi.org/10.1016/j.simpa.2023.100525
Vancouver
Meshram V, Suryawanshi Y, Meshram V, Patil K. Addressing misclassification in deep learning: A Merged Net approach. Software Impacts. 2023;17:100525. doi:10.1016/j.simpa.2023.100525
BibTeX
@article{Meshram2023,
author = {Meshram, Vishal and Suryawanshi, Yogesh and Meshram, Vidula and Patil, Kailas},
title = {Addressing misclassification in deep learning: A Merged Net approach},
journal = {Software Impacts},
year = {2023},
volume = {17},
pages = {100525},
doi = {10.1016/j.simpa.2023.100525}
}
EUVIMED