A DenseNet-121 model with CBAM attention, weighted cross-entropy, and CutMix achieves 96.90% accuracy on an eight-class onion disease image dataset.
Imbalanced Dataset Sampler,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Handling imbalance and few-sample size in ML based Onion disease classification
A DenseNet-121 model with CBAM attention, weighted cross-entropy, and CutMix achieves 96.90% accuracy on an eight-class onion disease image dataset.