CVCLMar 1, 2024

Multimodal ArXiv: A Dataset for Improving Scientific Comprehension of Large Vision-Language Models

Peking U
arXiv:2403.00231v3135 citationsh-index: 29Has CodeACL
Originality Synthesis-oriented
AI Analysis

This addresses the challenge of interpreting abstract scientific figures for researchers and developers in AI, though it is incremental as it builds on existing LVLM capabilities with new data.

The authors tackled the problem of limited scientific comprehension in large vision-language models (LVLMs) by introducing Multimodal ArXiv, a dataset comprising ArXivCap and ArXivQA, which improved LVLMs' mathematical reasoning by achieving a 10.4% absolute accuracy gain on a benchmark.

Large vision-language models (LVLMs) excel across diverse tasks involving concrete images from natural scenes. However, their ability to interpret abstract figures, such as geometry shapes and scientific plots, remains limited due to a scarcity of training datasets in scientific domains. To fill this gap, we introduce Multimodal ArXiv, consisting of ArXivCap and ArXivQA, for enhancing LVLMs scientific comprehension. ArXivCap is a figure-caption dataset comprising 6.4M images and 3.9M captions, sourced from 572K ArXiv papers spanning various scientific domains. Drawing from ArXivCap, we introduce ArXivQA, a question-answering dataset generated by prompting GPT-4V based on scientific figures. ArXivQA greatly enhances open-sourced LVLMs' mathematical reasoning capabilities, achieving a 10.4\% absolute accuracy gain on a multimodal mathematical reasoning benchmark. Furthermore, employing ArXivCap, we devise four vision-to-text tasks for benchmarking LVLMs. Evaluation results with state-of-the-art LVLMs underscore their struggle with the nuanced semantics of academic figures, while domain-specific training yields substantial performance gains. Our error analysis uncovers misinterpretations of visual context, recognition errors, and the production of overly simplified captions by current LVLMs, shedding light on future improvements.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes