arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1904.01906cs.CV
收藏

场景文本识别模型比较有什么问题?数据集和模型分析

What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis

Clova AI Research, NAVER/LINE Corp.(NAVER/LINE公司Clova AI研究中心);Kyoto University(京都大学)

Jeonghun Baek, Geewook Kim, Junyeop Lee, Sungrae Park, Dongyoon Han, Sangdoo Yun, Seong Joon Oh, Hwalsuk Lee

摘要:

近年来,出现了许多新的场景文本识别(STR)模型。虽然每个人都声称已经推动了技术的边界,但由于培训和评估数据集的选择不一致,该领域在很大程度上缺乏全面和公平的比较。本文解决了这一困难与三个主要贡献。首先,我们检查训练和评估数据集的不一致性,以及不一致性导致的性能差距。其次,我们介绍了一个统一的四阶段STR框架,大多数现有的STR模型适合。使用这个框架允许以前提出的STR模块和以前未开发的模块组合的发现进行广泛的评估。第三,我们分析了在一组一致的训练和评估数据集下,模块对准确性,速度和内存需求方面的性能贡献。这种分析清除了当前比较的障碍,以了解现有模块的性能增益。

英文摘要:

Many new proposals for scene text recognition (STR) models have been introduced in recent years. While each claim to have pushed the boundary of the technology, a holistic and fair comparison has been largely missing in the field due to the inconsistent choices of training and evaluation datasets. This paper addresses this difficulty with three major contributions. First, we examine the inconsistencies of training and evaluation datasets, and the performance gap results from inconsistencies. Second, we introduce a unified four-stage STR framework that most existing STR models fit into. Using this framework allows for the extensive evaluation of previously proposed STR modules and the discovery of previously unexplored module combinations. Third, we analyze the module-wise contributions to performance in terms of accuracy, speed, and memory demand, under one consistent set of training and evaluation datasets. Such analyses clean up the hindrance on the current comparisons to understand the performance gain of the existing modules.

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。

↑