TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Huayang, Li, Siheng, Cai, Deng, Wang, Longyue, Liu, Lemao, Watanabe, Taro, Yang, Yujiu, Shi, Shuming
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913374617468928
author Li, Huayang
Li, Siheng
Cai, Deng
Wang, Longyue
Liu, Lemao
Watanabe, Taro
Yang, Yujiu
Shi, Shuming
author_facet Li, Huayang
Li, Siheng
Cai, Deng
Wang, Longyue
Liu, Lemao
Watanabe, Taro
Yang, Yujiu
Shi, Shuming
contents Large language models with instruction-following abilities have revolutionized the field of artificial intelligence. These models show exceptional generalizability to tackle various real-world tasks through their natural language interfaces. However, their performance heavily relies on high-quality exemplar data, which is often difficult to obtain. This challenge is further exacerbated when it comes to multimodal instruction following. We introduce TextBind, an almost annotation-free framework for empowering larger language models with the multi-turn interleaved multimodal instruction-following capabilities. Our approach requires only image-caption pairs and generates multi-turn multimodal instruction-response conversations from a language model. To accommodate interleaved image-text inputs and outputs, we devise MIM, a language model-centric architecture that seamlessly integrates image encoder and decoder models. We release our dataset, model, and demo to foster future research in the area of multimodal instruction following.
format Preprint
id arxiv_https___arxiv_org_abs_2309_08637
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild
Li, Huayang
Li, Siheng
Cai, Deng
Wang, Longyue
Liu, Lemao
Watanabe, Taro
Yang, Yujiu
Shi, Shuming
Computation and Language
Artificial Intelligence
Large language models with instruction-following abilities have revolutionized the field of artificial intelligence. These models show exceptional generalizability to tackle various real-world tasks through their natural language interfaces. However, their performance heavily relies on high-quality exemplar data, which is often difficult to obtain. This challenge is further exacerbated when it comes to multimodal instruction following. We introduce TextBind, an almost annotation-free framework for empowering larger language models with the multi-turn interleaved multimodal instruction-following capabilities. Our approach requires only image-caption pairs and generates multi-turn multimodal instruction-response conversations from a language model. To accommodate interleaved image-text inputs and outputs, we devise MIM, a language model-centric architecture that seamlessly integrates image encoder and decoder models. We release our dataset, model, and demo to foster future research in the area of multimodal instruction following.
title TextBind: Multi-turn Interleaved Multimodal Instruction-following in the Wild
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2309.08637