Empirical comparison finds CNNs more robust on small datasets with local textures while ViTs perform better on complex global scenes but require more data and compute.
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
UNVERDICTED 1representative citing papers
citing papers explorer
No citing papers match the current filters.