Small-Large Collaboration: Training-efficient Concept Personalization for Large VLM using a Meta Personalized Small VLM

TL;DR

This paper introduces a training-efficient collaborative framework called Small-Large Collaboration (SLC) that personalizes large vision-language models by leveraging a meta personalized small VLM, combining the strengths of both models for improved reasoning and personalization.

Contribution

The paper proposes the first training-efficient framework for personalizing large VLMs using a meta personalized small VLM, applicable to both open-source and closed-source models.

Findings

SLC effectively personalizes large VLMs across various benchmarks.

The framework reduces training costs by only training a small VLM.

Experimental results show improved reasoning and personalization capabilities.

Abstract

Personalizing Vision-Language Models (VLMs) to transform them into daily assistants has emerged as a trending research direction. However, leading companies like OpenAI continue to increase model size and develop complex designs such as the chain of thought (CoT). While large VLMs are proficient in complex multi-modal understanding, their high training costs and limited access via paid APIs restrict direct personalization. Conversely, small VLMs are easily personalized and freely available, but they lack sufficient reasoning capabilities. Inspired by this, we propose a novel collaborative framework named Small-Large Collaboration (SLC) for large VLM personalization, where the small VLM is responsible for generating personalized information, while the large model integrates this personalized information to deliver accurate responses. To effectively incorporate personalized information, we develop a test-time reflection strategy, preventing the potential hallucination of the small VLM. Since SLC only needs to train a meta personalized small VLM for the large VLMs, the overall process is training-efficient. To the best of our knowledge, this is the first training-efficient framework that supports both open-source and closed-source large VLMs, enabling broader real-world personalized applications. We conduct thorough experiments across various benchmarks and large VLMs to demonstrate the effectiveness of the proposed SLC framework. The code will be released at https://github.com/Hhankyangg/SLC.