得益于嵌入式超敏织物基应变传感技术，研究人员表示该可穿戴“智能” 语音识别项圈有望开创无声语音接口（SSI）领域新纪元。

A wearable ‘smart’ choker for speech recognition has the potential to redefine the field of silent speech interface (SSI), say researchers – thanks to embedded ultrasensitive textile strain sensor technology.

Where verbal communication is hindered, such as in locations with lots of background noise or where an individual has an existing speech impairment, SSI systems are a cutting-edge solution, enabling verbal communication without vocalisation. As such, it is a type of electronic lip-reading using human-computer interaction.

这款佩戴在脖子上的智能项圈能够捕捉喉咙的细微活动，被应变传感器采集作为电子信号输入到类脑电脑软件模型中进行处理和语言识别。它甚至可以检测到默读的文字并播报出来，这能帮助诸如因喉部手术而无法说话的人群。

Worn around the neck, the smart choker captures micromovements in the throat, which are then picked up by the strain sensor as an electrical signal and fed into brain-inspired computer software models for processing and speech recognition. It can pick up even silently mouthed words and broadcast it, which could help someone who is unable to speak following laryngeal surgery, for example.

该智能项圈的独特结构特征是石墨烯涂层纺织物上规整的贯通裂缝。这种石墨烯层的结构能够显著提升应变传感器的灵敏度。它可以动态处理喉部的细微活动，捕捉信息丰富的语言信号。然后，这些信号会被高效计算的神经网络处理，语言解码的处理精准度高达95.25%。

The smart choker’s unique structure features ordered thorough cracks on graphene-coated textiles. The structured graphene layer significantly enhances the sensitivity of the strain sensor. It can dynamically respond to throat micromovements, enabling the capture of information-rich speech signals. These signals are then processed through a computationally efficient neural network, with a record accuracy of 95.25% in speech decoding. 

研究结果发表在《Npj-柔性电子》期刊上，该研究为实用的可穿戴SSI系统提供了一种非侵入式的解决方案，为在各种环境中实现无缝、自然的无声交流开辟了道路。

The results, reported in the journal npj Flexible Electronics, offer a promising, non-invasive solution for practical, wearable SSI systems, paving the way for seamless, natural silent communication in diverse settings.

该SSI系统性能强大，能够解码大量文字，并且可以快速地适应新用户和词汇。这项技术近期在IEEE2024年度生物传感器会议上做了现场展示，吸引了180多名与会者的注意。

The proposed SSI system is robust and able to decode a wide range of words, while swiftly adapting to new users and vocabularies. The technology was recently showcased as a live demonstration at the IEEE Biosensors 2024 conference, attracting the attention of 180 plus participants to the event.

该图为可穿戴SSI设备概览，主要靠超敏变应传感器和神经网络发挥作用，可用于高效的语音识别。

An overview of the wearable SSI, featuring an ultrasensitive strain sensor and a neural network for efficient speech recognition.

该研究由智能电子、生物系统与人工智能研究组组长兼剑桥大学工程系电气工程部和剑桥石墨烯中心下属Occhipinti研究小组负责人，Luigi G. Occhipinti博士领导。

The research was led by Dr Luigi G. Occhipinti, Director of Research in Smart Electronics, Biosystems and AI, and Head of the Occhipinti Group in the Electrical Engineering Division at the Department of Engineering and the Cambridge Graphene Centre.

Occhipinti博士表示：“我们的用户友好型智能项圈在现实场景中展现出了卓越的性能，它能够高效地适应不同性别、地域和种族背景的用户，处理长度和熟悉度各异的新词和可能含糊不清的词汇，并适应不同的语速。”

Dr Occhipinti said: “Our user-friendly smart choker demonstrates a remarkable ability to efficiently perform in real-world scenarios with users of different genders, geographical, and ethnic backgrounds, across new and potentially ambiguous words of varying lengths and familiarity, and at varying reading speeds.

“我们的SSI系统在区分不同用户的语音时具有高精度和计算效率，能够应对由传感器缺陷、外部环境以及用户佩戴设备时自身身体运动引起的各种噪声。”

“Our SSI system operates with high precision and computational efficiency in distinguishing the speech of different users against different types of noise, caused by sensor imperfections, the external environment, and/or from the users’ own body movements while wearing the device.

“此外，我们的超敏纺织基应变传感器技术的制造方法具有生物相容性、简单、低成本及可扩展性。该设备还适合于长期使用，能够经受超过10,000次的拉伸-释放循环，同时保持稳定可靠的电性功能。”

“Furthermore, the fabrication method of our ultrasensitive textile strain sensor technology is biocompatible, simple, low cost, and scalable. It is also adaptable to prolonged use and can withstand more than 10,000 stretching-releasing cycles while maintaining stable and reliable electrical functionality.