Saved in:
Bibliographic Details
Main Authors: Doveh, Sivan, Perek, Shaked, Mirza, M. Jehanzeb, Lin, Wei, Alfassy, Amit, Arbelle, Assaf, Ullman, Shimon, Karlinsky, Leonid
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.12736
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910529953464320
author Doveh, Sivan
Perek, Shaked
Mirza, M. Jehanzeb
Lin, Wei
Alfassy, Amit
Arbelle, Assaf
Ullman, Shimon
Karlinsky, Leonid
author_facet Doveh, Sivan
Perek, Shaked
Mirza, M. Jehanzeb
Lin, Wei
Alfassy, Amit
Arbelle, Assaf
Ullman, Shimon
Karlinsky, Leonid
contents State-of-the-art Vision-Language Models (VLMs) ground the vision and the language modality primarily via projecting the vision tokens from the encoder to language-like tokens, which are directly fed to the Large Language Model (LLM) decoder. While these models have shown unprecedented performance in many downstream zero-shot tasks (eg image captioning, question answers, etc), still little emphasis has been put on transferring one of the core LLM capability of In-Context Learning (ICL). ICL is the ability of a model to reason about a downstream task with a few examples demonstrations embedded in the prompt. In this work, through extensive evaluations, we find that the state-of-the-art VLMs somewhat lack the ability to follow ICL instructions. In particular, we discover that even models that underwent large-scale mixed modality pre-training and were implicitly guided to make use of interleaved image and text information (intended to consume helpful context from multiple images) under-perform when prompted with few-shot demonstrations (in an ICL way), likely due to their lack of direct ICL instruction tuning. To enhance the ICL abilities of the present VLM, we propose a simple yet surprisingly effective multi-turn curriculum-based learning methodology with effective data mixes, leading up to a significant 21.03% (and 11.3% on average) ICL performance boost over the strongest VLM baselines and a variety of ICL benchmarks. Furthermore, we also contribute new benchmarks for ICL evaluation in VLMs and discuss their advantages over the prior art.
format Preprint
id arxiv_https___arxiv_org_abs_2403_12736
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Multimodal In-Context Learning for Vision & Language Models
Doveh, Sivan
Perek, Shaked
Mirza, M. Jehanzeb
Lin, Wei
Alfassy, Amit
Arbelle, Assaf
Ullman, Shimon
Karlinsky, Leonid
Computer Vision and Pattern Recognition
State-of-the-art Vision-Language Models (VLMs) ground the vision and the language modality primarily via projecting the vision tokens from the encoder to language-like tokens, which are directly fed to the Large Language Model (LLM) decoder. While these models have shown unprecedented performance in many downstream zero-shot tasks (eg image captioning, question answers, etc), still little emphasis has been put on transferring one of the core LLM capability of In-Context Learning (ICL). ICL is the ability of a model to reason about a downstream task with a few examples demonstrations embedded in the prompt. In this work, through extensive evaluations, we find that the state-of-the-art VLMs somewhat lack the ability to follow ICL instructions. In particular, we discover that even models that underwent large-scale mixed modality pre-training and were implicitly guided to make use of interleaved image and text information (intended to consume helpful context from multiple images) under-perform when prompted with few-shot demonstrations (in an ICL way), likely due to their lack of direct ICL instruction tuning. To enhance the ICL abilities of the present VLM, we propose a simple yet surprisingly effective multi-turn curriculum-based learning methodology with effective data mixes, leading up to a significant 21.03% (and 11.3% on average) ICL performance boost over the strongest VLM baselines and a variety of ICL benchmarks. Furthermore, we also contribute new benchmarks for ICL evaluation in VLMs and discuss their advantages over the prior art.
title Towards Multimodal In-Context Learning for Vision & Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.12736