Abstract
The interaction between humans and artificial intelligence (AI) is a crucial factor that reflects the effectiveness of multimodal large language models (MLLMs). However, current MLLMs primarily focus on image-level comprehension and limit interaction to textual instructions, thereby constraining their flexibility in usage and depth of response. In this paper, we introduce the Draw-and-Understand project: a new model, a multi-domain dataset, and a challenging benchmark for visual prompting. Specifically, we propose SPHINX-V, a new end-to-end trained Multimodal Large Language Model (MLLM) that connects a vision encoder, a visual prompt encoder and an LLM for various visual prompts (points, bounding boxes, and free-form shape) and language understanding. To advance visual prompting research for MLLMs, we introduce MDVP-Data and MDVP-Bench. MDVP-Data features a multi-domain dataset containing 1.6M unique image-visual prompt-text instruction-following samples, including natural images, document images, OCR images, mobile screenshots, web screenshots, and multi-panel images. Furthermore, we present MDVP-Bench, a comprehensive and challenging benchmark to assess a model's capability in understanding visual prompting instructions. Our experiments demonstrate SPHINX-V's impressive multimodal interaction capabilities through visual prompting, revealing significant improvements in detailed pixel-level description and question-answering abilities.
Abstract (translated)
人类与人工智能(AI)之间的互动是一个反映多模态大型语言模型(MLLM)有效性的关键因素。然而,目前的MLLM主要关注图像级的理解,并将交互限制为文本指令,从而限制了其在使用和响应深度上的灵活性。在本文中,我们介绍了Draw-and-Understand项目:一个新的模型、一个多领域数据集和一个挑战性的基准,用于视觉提示。具体来说,我们提出了SPHINX-V,一种新的端到端训练的Multimodal Large Language Model(MLLM),它连接了视觉编码器、视觉提示编码器和各种视觉提示(点、边界框和自由形状)以及语言理解。为了促进MLLM在视觉提示方面的研究,我们引入了MDVP-Data和MDVP-Bench。MDVP-Data是一个多领域数据集,包含1600万独特图像-视觉提示-文本指令样本,包括自然图像、文档图像、OCR图像、移动屏幕截图、网页屏幕截图和多面板图像。此外,我们还介绍了MDVP-Bench,这是一个全面而具有挑战性的基准,用于评估模型对视觉提示指令的理解能力。我们的实验结果表明,通过视觉提示,SPHINX-V的令人印象深刻的多模态交互能力得到了展现,揭示了在详细像素级别描述和问题回答能力方面的显著改进。
URL
https://arxiv.org/abs/2403.20271