arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2302.01540cs.CV
收藏

DEVICE:用于基于OCR的图像描述的深度与视觉概念感知Transformer

DEVICE: Depth and Visual Concepts Aware Transformer for OCR-based Image Captioning

School of Artificial Intelligence at Guangxi University(广西大学人工智能学院);College of General Education, Guangxi Arts University(广西艺术学院文学院);The State Key Laboratory of Media Convergence and Communication, Communication University of China(中国传媒大学媒体融合与传播国家重点实验室);Guangxi Key Laboratory of Intelligent Control and Maintenance of Power Equipment(广西智能控制与电力设备关键实验室);School of Software Engineering, South China University of Technology(华南理工大学软件学院)

Dongsheng Xu, Qingbao Huang, Xingmao Zhang, Haonan Cheng, Feng Shuang, Yi Cai

AI总结:

针对基于OCR的图像描述中场景理解不足和描述不准确的问题,提出DEVICE模型,通过引入深度信息构建三维几何关系并利用视觉概念语义特征作为辅助信息,显著提升了描述的准确性和全面性。

摘要:

基于OCR的图像描述是一项重要但探索不足的任务,旨在生成包含视觉对象和场景文本的描述。最近的研究取得了令人鼓舞的进展,但仍然缺乏对场景的整体理解并生成不准确的描述。一个可能的原因是,当前的研究主要集中在构建场景文本的平面级几何关系,而没有深度信息。这导致场景文本关系推理不足,使得模型可能不准确描述场景文本。另一个可能的原因是,现有方法无法生成某些视觉对象的细粒度描述。此外,它们可能忽略基本的视觉对象,导致属于这些被忽略对象的场景文本未被利用。为了解决上述问题,我们提出了一种用于基于OCR的图像描述的深度与视觉概念感知Transformer(DEVICE)。具体而言,为了构建三维几何关系,我们引入深度信息并提出一个深度增强特征更新模块来改善OCR token特征。为了生成更精确和全面的描述,我们引入检测到的视觉概念的语义特征作为辅助信息,并提出一个语义引导对齐模块来提高模型利用视觉概念的能力。我们的DEVICE能够更全面地理解场景并提高描述视觉实体的准确性。充分的实验证明了我们提出的DEVICE的有效性,它在TextCaps测试集上优于最先进的模型。

英文摘要:

OCR-based image captioning is an important but under-explored task, aiming to generate descriptions containing visual objects and scene text. Recent studies have made encouraging progress, but they are still suffering from a lack of overall understanding of scenes and generating inaccurate captions. One possible reason is that current studies mainly focus on constructing the plane-level geometric relationship of scene text without depth information. This leads to insufficient scene text relational reasoning so that models may describe scene text inaccurately. The other possible reason is that existing methods fail to generate fine-grained descriptions of some visual objects. In addition, they may ignore essential visual objects, leading to the scene text belonging to these ignored objects not being utilized. To address the above issues, we propose a Depth and Visual Concepts Aware Transformer (DEVICE) for OCR-based image captinong. Concretely, to construct three-dimensional geometric relations, we introduce depth information and propose a depth-enhanced feature updating module to ameliorate OCR token features. To generate more precise and comprehensive captions, we introduce semantic features of detected visual concepts as auxiliary information, and propose a semantic-guided alignment module to improve the model's ability to utilize visual concepts. Our DEVICE is capable of comprehending scenes more comprehensively and boosting the accuracy of described visual entities. Sufficient experiments demonstrate the effectiveness of our proposed DEVICE, which outperforms state-of-the-art models on the TextCaps test set.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑