通过阅读顺序估计和动态采样检测反向类敌对场景文本
概括
这项研究引入了一种新的框架,用于发现反向类似场景文本,提高复杂和一般文本的准确性. 该方法有效地处理镜像和对称文本,而不会影响标准文本的性能.
科学领域:
- 计算机视觉 计算机视觉
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 发现场景文本是具有挑战性的,特别是对于复杂布局 (镜像,对称) 的反向类文本.
- 现有的方法与这些不规则的文本定向作斗争.
研究的目的:
- 提出一个统一的,端到端可训练框架 (IATS) 来发现反向类似场景文本.
- 有效地处理反向类文本,而不会降低对一般场景文本的性能.
主要方法:
- 引入了一种创新的阅读顺序估计模块 (REM),使用关节损失 (LRE).
- 使用初始边界模块 (IBM) 和边界精制模块 (BRM) 进行自适应文本边界检测.
- 开发了一个具有薄板支线的动态采样模块 (DSM),用于改进文本识别功能采样.
主要成果:
- 在具有挑战性的场景文本和反向类似场景文本数据集上实现了卓越的性能.
- 在高精度发现不规则和反向类文本方面表现出有效性.
- 在没有额外的监督的情况下,DSM主动学习了识别的最佳特征.
结论:
- 拟议的IATS框架为逆相似场景文本发现提供了一个强大的解决方案.
- 该方法可以很好地泛化到各种文本形状,尺度和方向.
- 这项工作提升了复杂视觉环境中的场景文本识别系统的能力.
更多相关视频
07:36Eye Tracking During Visually Situated Language Comprehension: Flexibility and Limitations in Uncovering Visual Context Effects
Published on: November 30, 2018
15.7K
09:09Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
Published on: September 27, 2024
459
相关概念视频
03:26Framing Effects
7.4K
Information is everywhere and its presentation—such as how and when items are presented—can impact our perceptions and decisions surrounding the info. This broad concept umbrellas framing effects—influences that occur due to the way information is framed in its appearance, whether it’s purely the order or the specific wording of a message. Let’s take a look at numerous ways in which two versions of something can objectively say the same thing, yet we respond in...
7.4K
02:01Group Design
8.9K
The most basic experimental design involves two groups: the experimental group and the control group. The two groups are designed to be the same except for one difference— experimental manipulation. The experimental group gets the experimental manipulation—that is, the treatment or variable being tested—and the control group does not. Since experimental manipulation is the only difference between the experimental and control groups, we can be sure that any differences between...
8.9K