A spectral vision transformer achieves equitable or superior performance with fewer parameters than standard ViTs, CNNs, and other models by using spectral projections for tokenization in limited-data medical imaging.
arXiv:2205.09723 (2022)
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2roles
background 2representative citing papers
Pre-training on modality-matched data significantly improves downstream performance in medical imaging models while self-supervised learning benefits depend on context.
citing papers explorer
-
Spectral Vision Transformer for Efficient Tokenization with Limited Data
A spectral vision transformer achieves equitable or superior performance with fewer parameters than standard ViTs, CNNs, and other models by using spectral projections for tokenization in limited-data medical imaging.
-
From pre-training to downstream performance: Does domain-specific pre-training make sense?
Pre-training on modality-matched data significantly improves downstream performance in medical imaging models while self-supervised learning benefits depend on context.