MiniGPT-v2: large language model as a unified interface for vision-language multi-task learning

TL;DR

MiniGPT-v2 is a unified vision-language model that uses task identifiers to effectively perform multiple tasks like image captioning, VQA, and grounding, achieving strong benchmark results.

Contribution

The paper introduces MiniGPT-v2, a multi-task vision-language model that employs unique task identifiers to improve task differentiation and learning efficiency.

Findings

Achieves strong performance on visual question-answering benchmarks.

Effectively handles multiple vision-language tasks with a single model.

Uses task identifiers to enhance task learning and differentiation.

Abstract

Large language models have shown their remarkable capabilities as a general interface for various language-related applications. Motivated by this, we target to build a unified interface for completing many vision-language tasks including image description, visual question answering, and visual grounding, among others. The challenge is to use a single model for performing diverse vision-language tasks effectively with simple multi-modal instructions. Towards this objective, we introduce MiniGPT-v2, a model that can be treated as a unified interface for better handling various vision-language tasks. We propose using unique identifiers for different tasks when training the model. These identifiers enable our model to better distinguish each task instruction effortlessly and also improve the model learning efficiency for each task. After the three-stage training, the experimental results show that MiniGPT-v2 achieves strong performance on many visual question-answering and visual grounding benchmarks compared to other vision-language generalist models. Our model and codes are available at https://minigpt-v2.github.io/