EfficientNetB0 with Geometric & Color augmentation and TrivialAugmentWide achieves the best performance for rabies classification on a small imbalanced set of 155 fluorescent images via 3-fold cross-validation.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Rabies diagnosis in low-data settings: A comparative study on the impact of data augmentation and transfer learning
EfficientNetB0 with Geometric & Color augmentation and TrivialAugmentWide achieves the best performance for rabies classification on a small imbalanced set of 155 fluorescent images via 3-fold cross-validation.