Progressive Vision-Language Alignment via LLM-based Embedder

Xiaoxing Hu, Kaicheng Yang, Ziyang Gong, Qi Ming, Zonghao Guo, Xiang An, Ziyong Feng, Junchi Yan, Xue Yang

Keywords: vision language representation, multi-modality, CLIP

TL;DR: An alignment framework between CLIP Image encoder and LLM-based embedder

Abstract: The original CLIP text encoder is limited by a maximum input length of 77 tokens, which hampers its ability to effectively process long texts and perform fine-grained semantic understanding. In addition, the CLIP text encoder lacks support for multilingual inputs. All these limitations significantly restrict its applicability across a broader range of tasks. Recent studies have attempted to replace the CLIP text encoder with an LLM-based embedder to enhance its ability in processing long texts, multilingual understanding, and fine-grained semantic comprehension. However, because the representation spaces of LLMs and the vision-language space of CLIP are pretrained independently without alignment priors, direct alignment using contrastive learning can disrupt the intrinsic vision-language alignment in the CLIP image encoder, leading to an underutilization of the knowledge acquired during pre-training. To address this challenge, we propose ProCLIP, a curriculum learning-based progressive vision-language alignment framework to effectively align the CLIP image encoder with an LLM-based embedder. Specifically, ProCLIP first distills knowledge from CLIP’s text encoder into the LLM-based embedder to leverage CLIP's rich pretrained knowledge while establishing initial alignment between the LLM embedder and CLIP image encoder. Subsequently, ProCLIP further aligns the CLIP image encoder with the LLM-based embedder through image-text contrastive tuning, employing self-distillation regularization to avoid overfitting. To achieve a more effective alignment, instance semantic alignment loss and embedding structure alignment loss are employed during representation inheritance and contrastive tuning. Extensive experiments show ProCLIP achieves 6.8% to 13.5% improvement on zero-shot classification and presents excellent performance on cross-modal retrieval, multilingual cross-modal retrieval, and fine-grained understanding tasks, demonstrating the effectiveness and robustness of ProCLIP. To support reproducibility and facilitate further research, we will release the training code and model weights.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 3789