A Multiscale Fusion Network Integrating ConvNeXtSmall and EfficientNetB0 with Enhanced Image Preprocessing for Robust Classification

Authors

  • Sanjay Balwani Department of Electronics and Telecommunication Engineering, Jhulelal Institute of Technology, India
  • Narendra Bawane Department of Electronics and Telecommunication Engineering, Jhulelal Institute of Technology, India

DOI:

https://doi.org/10.63278/1388

Keywords:

Multiscale Fusion Network, ConvNeXtSmall, EfficientNetB0, Multiscale Retinex, GrabCut, Image Classification, Preprocessing Techniques, Deep Learning.

Abstract

In this paper, we introduce the Multiscale Fusion Network (MFNet), a pioneering approach that combines ConvNeXtSmall and EfficientNetB0 architectures with advanced preprocessing techniques to revolutionize image classification. By integrating Multiscale Retinex and GrabCut segmentation, MFNet enhances perceptual quality, extracts meaningful features, and minimizes background noise. Trained on a meticulously prepared dataset, our model demonstrates superior performance, achieving high accuracy and robustness across diverse datasets. This paper delves into the architectural synergy, preprocessing innovations, and rigorous evaluation that establish MFNet as a leading solution for image classification tasks.

Downloads

How to Cite

Sanjay Balwani, and Narendra Bawane. 2025. “A Multiscale Fusion Network Integrating ConvNeXtSmall and EfficientNetB0 With Enhanced Image Preprocessing for Robust Classification”. Metallurgical and Materials Engineering 31 (3):382-91. https://doi.org/10.63278/1388.

Issue

Vol. 31 No. 3 (2025)

Section

Research

License

Copyright (c) 2025 Sanjay Balwani, Narendra Bawane

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Authors who publish with this journal agree to the following terms:

  • Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
  • Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
  • Authors are permitted and encouraged to post their published articles online (e.g., in institutional repositories or on their website, social networks like ResearchGate or Academia), as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).

 

Creative Commons License
Except where otherwise noted, the content on this site is licensed under a Creative Commons Attribution 4.0 International License.