Comparison of EfficientNetV2-S and ConvNeXt V2-Tiny for Potato Leaf Condition Classification with Grad-CAM
Asian Journal of Social and Humanities · 25.01.2022 · 1 (01) · 40-56
DOI: 10.59888/ajosh.v1i01.786 ↗Kurz gefasst
Fachartikel in der Zeitschrift Asian Journal of Social and Humanities, online erschienen am 25.01.2022. Verfasst von Yoga Pamungkas. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Background
Variations in background, lighting, and viewing angle make potato leaf condition classification difficult in uncontrolled environments.
Objective
This study compared EfficientNetV2-S and ConvNeXt V2-Tiny on seven classes in terms of classification performance and inference efficiency and examined the image regions contributing to predictions using Gradient-weighted Class Activation Mapping (Grad-CAM).
Method
Of 3,076 initial images, 12 with label conflicts were quarantined and 49 identical copies removed, leaving 3,015 images, split into 2,412 training, 301 validation, and 302 test images with similar images kept in the same partition. Both models were trained using transfer learning, weighted cross-entropy, and three seeds.
Results
EfficientNetV2-S achieved a mean test accuracy of 83.22% and macro-F1 of 79.86%, whereas ConvNeXt V2-Tiny achieved 87.97% and 85.21%. The macro-F1 difference of 5.34 percentage points had a 95% confidence interval of 2.43–8.50 percentage points based on a paired group bootstrap, although the per-run advantage ranged from 1.06 to 11.06 points and the seed-42 McNemar test was not significant (p ? 0.39). On an NVIDIA T4 GPU, ConvNeXt V2-Tiny required 11.70 versus 25.39 ms per image. Grad-CAM visualizations showed different activation spreads for the two models on example test images.
Conclusion
ConvNeXt V2-Tiny delivered higher performance and lower latency in this internal test and is the more practical candidate for server-based diagnosis, although cross-location evaluation and spatial validation against lesion annotations are still required.
Zitieren
APA
Pamungkas, Y. (2022). Comparison of EfficientNetV2-S and ConvNeXt V2-Tiny for Potato Leaf Condition Classification with Grad-CAM. Asian Journal of Social and Humanities, 1(01), 40-56. https://doi.org/10.59888/ajosh.v1i01.786
Vancouver
Pamungkas Y. Comparison of EfficientNetV2-S and ConvNeXt V2-Tiny for Potato Leaf Condition Classification with Grad-CAM. Asian Journal of Social and Humanities. 2022;1(01):40-56. doi:10.59888/ajosh.v1i01.786
BibTeX
@article{Pamungkas2022,
author = {Pamungkas, Yoga},
title = {Comparison of EfficientNetV2-S and ConvNeXt V2-Tiny for Potato Leaf Condition Classification with Grad-CAM},
journal = {Asian Journal of Social and Humanities},
year = {2022},
volume = {1},
number = {01},
pages = {40--56},
doi = {10.59888/ajosh.v1i01.786}
}
EUVIMED