RLEG: Vision-Language Representation Learning with Diffusion-based Embedding Generation

Liming Zhao, Kecheng Zheng, Yun Zheng, Deli Zhao, Jingren Zhou
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:42247-42258, 2023.

Abstract

Vision-language representation learning models (e.g., CLIP) have achieved state-of-the-art performance on various downstream tasks, which usually need large-scale training data to learn discriminative representation. Recent progress on generative diffusion models (e.g., DALL-E 2) has demonstrated that diverse high-quality samples can be synthesized by randomly sampling from generative distribution. By virtue of generative capability in this paper, we propose a novel vision-language Representation Learning method with diffusion-based Embedding Generation (RLEG), which exploits diffusion models to generate feature embedding online for learning effective vision-language representation. Specifically, we first adopt image and text encoders to extract the corresponding embeddings. Secondly, pretrained diffusion-based embedding generators are harnessed to transfer the embedding modality online between vision and language domains. The embeddings generated from the generators are then served as augmented embedding-level samples, which are applied to contrastive learning with the variant of the CLIP framework. Experimental results show that the proposed method could learn effective representation and achieve state-of-the-art performance on various tasks including image classification, image-text retrieval, object detection, semantic segmentation, and text-conditional image generation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-zhao23l, title = {{RLEG}: Vision-Language Representation Learning with Diffusion-based Embedding Generation}, author = {Zhao, Liming and Zheng, Kecheng and Zheng, Yun and Zhao, Deli and Zhou, Jingren}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {42247--42258}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/zhao23l/zhao23l.pdf}, url = {https://proceedings.mlr.press/v202/zhao23l.html}, abstract = {Vision-language representation learning models (e.g., CLIP) have achieved state-of-the-art performance on various downstream tasks, which usually need large-scale training data to learn discriminative representation. Recent progress on generative diffusion models (e.g., DALL-E 2) has demonstrated that diverse high-quality samples can be synthesized by randomly sampling from generative distribution. By virtue of generative capability in this paper, we propose a novel vision-language Representation Learning method with diffusion-based Embedding Generation (RLEG), which exploits diffusion models to generate feature embedding online for learning effective vision-language representation. Specifically, we first adopt image and text encoders to extract the corresponding embeddings. Secondly, pretrained diffusion-based embedding generators are harnessed to transfer the embedding modality online between vision and language domains. The embeddings generated from the generators are then served as augmented embedding-level samples, which are applied to contrastive learning with the variant of the CLIP framework. Experimental results show that the proposed method could learn effective representation and achieve state-of-the-art performance on various tasks including image classification, image-text retrieval, object detection, semantic segmentation, and text-conditional image generation.} }
Endnote
%0 Conference Paper %T RLEG: Vision-Language Representation Learning with Diffusion-based Embedding Generation %A Liming Zhao %A Kecheng Zheng %A Yun Zheng %A Deli Zhao %A Jingren Zhou %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-zhao23l %I PMLR %P 42247--42258 %U https://proceedings.mlr.press/v202/zhao23l.html %V 202 %X Vision-language representation learning models (e.g., CLIP) have achieved state-of-the-art performance on various downstream tasks, which usually need large-scale training data to learn discriminative representation. Recent progress on generative diffusion models (e.g., DALL-E 2) has demonstrated that diverse high-quality samples can be synthesized by randomly sampling from generative distribution. By virtue of generative capability in this paper, we propose a novel vision-language Representation Learning method with diffusion-based Embedding Generation (RLEG), which exploits diffusion models to generate feature embedding online for learning effective vision-language representation. Specifically, we first adopt image and text encoders to extract the corresponding embeddings. Secondly, pretrained diffusion-based embedding generators are harnessed to transfer the embedding modality online between vision and language domains. The embeddings generated from the generators are then served as augmented embedding-level samples, which are applied to contrastive learning with the variant of the CLIP framework. Experimental results show that the proposed method could learn effective representation and achieve state-of-the-art performance on various tasks including image classification, image-text retrieval, object detection, semantic segmentation, and text-conditional image generation.
APA
Zhao, L., Zheng, K., Zheng, Y., Zhao, D. & Zhou, J.. (2023). RLEG: Vision-Language Representation Learning with Diffusion-based Embedding Generation. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:42247-42258 Available from https://proceedings.mlr.press/v202/zhao23l.html.

Related Material