MM-LG extracts a compact multimodal and unimodal block set from CLIP via distillation and uses it to initialize smaller vision-language and vision models, outperforming previous Learngene methods and sometimes pre-training plus fine-tuning.
Align before fuse: Vision and language representation learning with momentum distilla- tion.Advances in neural information processing systems, 34:9694–9705,
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
-
Extracting Multimodal Learngene in CLIP: Unveiling the Multimodal Generalizable Knowledge
MM-LG extracts a compact multimodal and unimodal block set from CLIP via distillation and uses it to initialize smaller vision-language and vision models, outperforming previous Learngene methods and sometimes pre-training plus fine-tuning.