DEFTS Offers Data-Efficient, Training-Free Few-Shot Learning for CLIP in Text Subspace
A new arXiv paper introduces DEFTS, a data-efficient, training-free few-shot learning method for CLIP that operates in the text subspace, building on CLIP's zero-shot image classification approach.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-18 05:00 · arXiv stat.ML
DEFTS: Data-Efficient, Training-Free Few-Shot Learning for CLIP in the Text Subspace