LoRA fine-tuning of pretrained vision models beats head-only fine-tuning on four cytology benchmarks, and CLIP with LoRA matches the HiCervix state of the art using 70% of the data.
For instance, 2 shots of the SIPaKMeD dataset results a total of 10 examples (two examples per class)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Exploring Foundation Models Fine-Tuning for Cytology Classification
LoRA fine-tuning of pretrained vision models beats head-only fine-tuning on four cytology benchmarks, and CLIP with LoRA matches the HiCervix state of the art using 70% of the data.