A Transformer with a loosely defined multi-scale attention weighting is reported to achieve 89.5% accuracy on ISIC 2017 skin lesion classification.
Skin lesion classification for melanoma using deep learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Multi-Scale Transformer Architecture for Accurate Medical Image Classification
A Transformer with a loosely defined multi-scale attention weighting is reported to achieve 89.5% accuracy on ISIC 2017 skin lesion classification.