Suppr超能文献

舌后部分辨率低是空间模式识别的瓶颈。

Poor resolution at the back of the tongue is the bottleneck for spatial pattern recognition.

机构信息

The Schepens Eye Research Institute of Mass. Eye and Ear, Department of Ophthalmology, Harvard Medical School, Boston, MA, USA.

出版信息

Sci Rep. 2020 Feb 12;10(1):2435. doi: 10.1038/s41598-020-59102-3.

Abstract

Spatial patterns presented on the tongue using electro-tactile sensory substitution devices (SSDs) have been suggested to be recognized better by tracing the pattern with the tip of the tongue. We examined if the functional benefit of tracing is overcoming the poor sensitivity or low spatial resolution at the back of the tongue or alternatively compensating for limited information processing capacity by fixating on a segment of the spatial pattern at a time. Using a commercially available SSD, the BrainPort, we compared letter recognition performance in three presentation modes; tracing, static, and drawing. Stimulation intensity was either constant or increased from the tip to the back of the tongue to partially compensate for the decreasing sensitivity. Recognition was significantly better for tracing, compared to static and drawing conditions. Confusion analyses showed that letters were confused based on their characteristics presented near the tip in static and drawing conditions. The results suggest that recognition performance is limited by the poor spatial resolution at the back of the tongue, and tracing seems to be an effective strategy to overcome this. Compensating for limited information processing capacity or poor sensitivity by drawing or increasing intensity at the back, respectively, does not improve the performance.

摘要

使用电触觉感觉替代设备 (SSD) 在舌头上呈现的空间模式,通过用舌尖追踪模式,被认为可以更好地识别。我们研究了追踪是否可以克服舌后部的低灵敏度或低空间分辨率的功能优势,或者通过一次固定在空间模式的一个片段来弥补有限的信息处理能力。使用一种商业上可用的 SSD,即 BrainPort,我们比较了三种呈现模式下的字母识别性能:追踪、静态和绘制。刺激强度要么是恒定的,要么从舌尖向后舌逐渐增加,以部分补偿灵敏度的下降。与静态和绘制条件相比,追踪的识别性能明显更好。混淆分析表明,在静态和绘制条件下,字母是根据其在舌尖附近呈现的特征混淆的。结果表明,识别性能受到舌后部低空间分辨率的限制,而追踪似乎是克服这一限制的有效策略。通过绘制或在后舌增加强度来分别弥补有限的信息处理能力或低灵敏度并不能提高性能。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5b66/7015888/82ff5931fdb8/41598_2020_59102_Fig1_HTML.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验