Decoupling images into foreground and background before aligning them with text improves CLIP prompt tuning on few-shot and generalization benchmarks.
Learning to prompt for vision-language models,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Decouple before Align: Visual Disentanglement Enhances Prompt Tuning
Decoupling images into foreground and background before aligning them with text improves CLIP prompt tuning on few-shot and generalization benchmarks.