CLAMP: Contrastive LAnguage Model Prompt-tuning
This work addresses the challenge of leveraging LLMs for visual tasks like image classification, which could benefit researchers and practitioners in multimodal AI, though it is incremental as it builds on existing contrastive learning methods.
The paper tackles the problem of adapting large language models (LLMs) to image classification, showing that fine-tuning LLMs with a contrastive objective achieves good performance, beating state-of-the-art multimodal LLMs by 13% and slightly outperforming custom text models while retaining generative abilities.
Large language models (LLMs) have emerged as powerful general-purpose interfaces for many machine learning problems. Recent work has adapted LLMs to generative visual tasks like image captioning, visual question answering, and visual chat, using a relatively small amount of instruction-tuning data. In this paper, we explore whether modern LLMs can also be adapted to classifying an image into a set of categories. First, we evaluate multimodal LLMs that are tuned for generative tasks on zero-shot image classification and find that their performance is far below that of specialized models like CLIP. We then propose an approach for light fine-tuning of LLMs using the same contrastive image-caption matching objective as CLIP. Our results show that LLMs can, indeed, achieve good image classification performance when adapted this way. Our approach beats state-of-the-art mLLMs by 13% and slightly outperforms contrastive learning with a custom text model, while also retaining the LLM's generative abilities. LLM initialization appears to particularly help classification in domains under-represented in the visual pre-training data.