DCLIP fine-tunes a CLIP student's image encoder to match a YOLO-region, bidirectional cross-attention teacher, improving retrieval while retaining most zero-shot accuracy.
Title resolution pending
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
-
Distill CLIP (DCLIP): Enhancing Image-Text Retrieval via Cross-Modal Transformer Distillation
DCLIP fine-tunes a CLIP student's image encoder to match a YOLO-region, bidirectional cross-attention teacher, improving retrieval while retaining most zero-shot accuracy.