TransferCVLM: Transferring Cross-Modal Knowledge for Vision-Language Modeling

Dongha Choi, Jung-jae Kim, Hyunju Lee


Abstract
Recent large vision-language multimodal models pre-trained with huge amount of image-text pairs show remarkable performances in downstream tasks. However, the multimodal pre-training has limitations in terms of resources and training time when it comes to obtaining new models that surpass existing models. To overcome these issues, we propose TransferCVLM, a method of efficient knowledge transfer that integrates pre-trained uni-modal models (and cross-modal fusion-encoder) into a combined vision-language model (CVLM), without pre-training the CVLM with large amount of multimodal data, and then for each task application, fine-tunes the CVLM and transfers the multimodal knowledge of a teacher vision-language model to the CVLM by using knowledge distillation techniques. We demonstrate that 1) the fine-tuned CVLM performs comparable to other vision-language models of similar size, that 2) the multimodal knowledge transfer consistently enhances the CVLM, and the knowledge-transferred CVLM composed of large-size unimodal models outperforms the teacher multimodal model in most of downstream tasks, and that 3) TransferCVLM can also be used for model compression when using small-size unimodal models. We estimate that the training of TransferCVLM takes only 6% of pre-training of other vision-language models. Our code is available at https://github.com/DMCB-GIST/TransferCVLM.
Anthology ID:
2024.findings-emnlp.975
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2024
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
16733–16746
Language:
URL:
https://aclanthology.org/2024.findings-emnlp.975
DOI:
Bibkey:
Cite (ACL):
Dongha Choi, Jung-jae Kim, and Hyunju Lee. 2024. TransferCVLM: Transferring Cross-Modal Knowledge for Vision-Language Modeling. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 16733–16746, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
TransferCVLM: Transferring Cross-Modal Knowledge for Vision-Language Modeling (Choi et al., Findings 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.findings-emnlp.975.pdf
Software:
 2024.findings-emnlp.975.software.zip