Improving the Input Accuracy of Touchscreens using Deep Learning

Improving the Input Accuracy of Touchscreens using Deep Learning
复制标题

使用深度学习提高触摸屏的输入精度

DOI:
10.1145/3290607.3312928
复制
发表时间:
2019
期刊:
Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems
影响因子:
--
通讯作者:
Huy Viet Le
Huy Viet Le
中科院分区:
--
文献类型:
--
作者:
Abinaya Kumar;Aishwarya Radjesh;Sven Mayer;Huy Viet Le

文献摘要

参考文献

被引文献

相似文献

触摸屏将联合收割机的输入和输出结合在一个界面中。虽然这实现了直观的交互和动态用户界面,但胖手指问题和由此产生的遮挡仍然影响输入准确性。先前的工作提出了通过涉及手指顶侧的视觉特征以及静态补偿功能来提高触摸精度的方法。虽然前者不适用于最近的移动的设备,因为手指的顶侧不能被跟踪,但是补偿功能不考虑诸如手指角度的属性。在这项工作中,我们提出了一种数据驱动的方法来估计商品互电容触摸屏上的2D触摸位置,该方法比最近实施的方法提高了23.0%的触摸精度。
Touchscreens combine input and output in a single interface. While this enables an intuitive interaction and dynamic user interfaces, the fat-finger problem and the resulting occlusions still impact the input accuracy. Previous work presented approaches to improve the touch accuracy by involving visual features on the top side of fingers, as well as static compensation functions. While the former is not applicable on recent mobile devices as the top side of a finger cannot be tracked, compensation functions do not take properties such as finger angle into account. In this work, we present a data-driven approach to estimate the 2D touch position on commodity mutual capacitive touchscreens which increases the touch accuracy by 23.0% over recently implemented approaches.
100,000,000次点击:大范围触摸性能分析与改进
DOI: 10.1145/2037373.2037395
发表时间: 2011
期刊:
影响因子: --
作者:
Niels Henze;Enrico Rukzio;Susanne Boll
通讯作者: Susanne Boll