PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
April 14, 20242 citationsOpen Access

TextHawk: Exploring Efficient Fine-Grained Perception of Multimodal Large Language Models

View Full Paper
YYYaqi YuMLMinghui LiaoJWJihao Wu

Key Points

Key points are not available for this paper at this time.

Abstract

Multimodal Large Language Models (MLLMs) have shown impressive results on various multimodal tasks. However, most existing MLLMs are not well suited for document-oriented tasks, which require fine-grained image perception and information compression. In this paper, we present TextHawk, a MLLM that is specifically designed for document-oriented tasks, while preserving the general capabilities of MLLMs. TextHawk is aimed to explore efficient fine-grained perception by designing four dedicated components. Firstly, a ReSampling and ReArrangement (ReSA) module is proposed to reduce the redundancy in the document texts and lower the computational cost of the MLLM. We explore encoding the positions of each local feature by presenting Scalable Positional Embeddings (SPEs), which can preserve the scalability of various image sizes. A Query Proposal Network (QPN) is then adopted to initialize the queries dynamically among different sub-images. To further enhance the fine-grained visual perceptual ability of the MLLM, we design a Multi-Level Cross-Attention (MLCA) mechanism that captures the hierarchical structure and semantic relations of document images. Furthermore, we create a new instruction-tuning dataset for document-oriented tasks by enriching the multimodal document data with Gemini Pro. We conduct extensive experiments on both general and document-oriented MLLM benchmarks, and show that TextHawk outperforms the state-of-the-art methods, demonstrating its effectiveness and superiority in fine-grained document perception and general abilities.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Yu et al. (2024) studied this question.

synapsesocial.com/papers/68e6f3b7b6db64358766eabbhttps://doi.org/10.48550/arxiv.2404.09204
Ask AI
Helpful
Bookmark
Share
View Full Paper