Multimodal Few-Shot Learning with Frozen Language Models

TL;DR

This paper introduces a simple method to extend the few-shot learning capabilities of large language models to multimodal tasks involving vision and language, using aligned image-caption data and a frozen language model.

Contribution

The authors propose a novel approach that trains a vision encoder to produce image embeddings compatible with a frozen language model for multimodal few-shot learning.

Findings

The system can learn new visual categories rapidly.

It performs visual question-answering with minimal examples.

It leverages outside knowledge effectively.

Abstract

When trained at sufficient scale, auto-regressive language models exhibit the notable ability to learn a new language task after being prompted with just a few examples. Here, we present a simple, yet effective, approach for transferring this few-shot learning ability to a multimodal setting (vision and language). Using aligned image and caption data, we train a vision encoder to represent each image as a sequence of continuous embeddings, such that a pre-trained, frozen language model prompted with this prefix generates the appropriate caption. The resulting system is a multimodal few-shot learner, with the surprising ability to learn a variety of new tasks when conditioned on examples, represented as a sequence of multiple interleaved image and text embeddings. We demonstrate that it can rapidly learn words for new objects and novel visual categories, do visual question-answering with only a handful of examples, and make use of outside knowledge, by measuring a single model on a variety of established and new benchmarks.