CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers

Dachuan Shi, Chaofan Tao, Anyi Rao, Zhendong Yang, Chun Yuan, Jiaqi Wang
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:44960-44990, 2024.

Abstract

Recent vision-language models have achieved tremendous advances. However, their computational costs are also escalating dramatically, making model acceleration exceedingly critical. To pursue more efficient vision-language Transformers, this paper introduces Cross-Guided Ensemble of Tokens (CrossGET), a general acceleration framework for vision-language Transformers. This framework adaptively combines tokens in real-time during inference, significantly reducing computational costs while maintaining high performance. CrossGET features two primary innovations: 1) Cross-Guided Matching and Ensemble. CrossGET leverages cross-modal guided token matching and ensemble to effectively utilize cross-modal information, achieving wider applicability across both modality-independent models, e.g., CLIP, and modality-dependent ones, e.g., BLIP2. 2) Complete-Graph Soft Matching. CrossGET introduces an algorithm for the token-matching mechanism, ensuring reliable matching results while facilitating parallelizability and high efficiency. Extensive experiments have been conducted on various vision-language tasks, such as image-text retrieval, visual reasoning, image captioning, and visual question answering. The performance on both classic multimodal architectures and emerging multimodal LLMs demonstrates the framework’s effectiveness and versatility. The code is available at https://github.com/sdc17/CrossGET.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-shi24e, title = {{C}ross{GET}: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers}, author = {Shi, Dachuan and Tao, Chaofan and Rao, Anyi and Yang, Zhendong and Yuan, Chun and Wang, Jiaqi}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {44960--44990}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/shi24e/shi24e.pdf}, url = {https://proceedings.mlr.press/v235/shi24e.html}, abstract = {Recent vision-language models have achieved tremendous advances. However, their computational costs are also escalating dramatically, making model acceleration exceedingly critical. To pursue more efficient vision-language Transformers, this paper introduces Cross-Guided Ensemble of Tokens (CrossGET), a general acceleration framework for vision-language Transformers. This framework adaptively combines tokens in real-time during inference, significantly reducing computational costs while maintaining high performance. CrossGET features two primary innovations: 1) Cross-Guided Matching and Ensemble. CrossGET leverages cross-modal guided token matching and ensemble to effectively utilize cross-modal information, achieving wider applicability across both modality-independent models, e.g., CLIP, and modality-dependent ones, e.g., BLIP2. 2) Complete-Graph Soft Matching. CrossGET introduces an algorithm for the token-matching mechanism, ensuring reliable matching results while facilitating parallelizability and high efficiency. Extensive experiments have been conducted on various vision-language tasks, such as image-text retrieval, visual reasoning, image captioning, and visual question answering. The performance on both classic multimodal architectures and emerging multimodal LLMs demonstrates the framework’s effectiveness and versatility. The code is available at https://github.com/sdc17/CrossGET.} }
Endnote
%0 Conference Paper %T CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers %A Dachuan Shi %A Chaofan Tao %A Anyi Rao %A Zhendong Yang %A Chun Yuan %A Jiaqi Wang %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-shi24e %I PMLR %P 44960--44990 %U https://proceedings.mlr.press/v235/shi24e.html %V 235 %X Recent vision-language models have achieved tremendous advances. However, their computational costs are also escalating dramatically, making model acceleration exceedingly critical. To pursue more efficient vision-language Transformers, this paper introduces Cross-Guided Ensemble of Tokens (CrossGET), a general acceleration framework for vision-language Transformers. This framework adaptively combines tokens in real-time during inference, significantly reducing computational costs while maintaining high performance. CrossGET features two primary innovations: 1) Cross-Guided Matching and Ensemble. CrossGET leverages cross-modal guided token matching and ensemble to effectively utilize cross-modal information, achieving wider applicability across both modality-independent models, e.g., CLIP, and modality-dependent ones, e.g., BLIP2. 2) Complete-Graph Soft Matching. CrossGET introduces an algorithm for the token-matching mechanism, ensuring reliable matching results while facilitating parallelizability and high efficiency. Extensive experiments have been conducted on various vision-language tasks, such as image-text retrieval, visual reasoning, image captioning, and visual question answering. The performance on both classic multimodal architectures and emerging multimodal LLMs demonstrates the framework’s effectiveness and versatility. The code is available at https://github.com/sdc17/CrossGET.
APA
Shi, D., Tao, C., Rao, A., Yang, Z., Yuan, C. & Wang, J.. (2024). CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:44960-44990 Available from https://proceedings.mlr.press/v235/shi24e.html.

Related Material