Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Silent Speech Interfaces in the Era of Large Language Models: A Comprehensive Taxonomy and Systematic Review

Kele Xu, Yifan Wang, Ming Feng, Qisheng Xu, Wuyang Chen, Yutao Dou, Cheng Yang, Huaimin Wang

Abstract

Human-computer interaction has traditionally relied on the acoustic channel, a dependency that introduces systemic vulnerabilities to environmental noise, privacy constraints, and physiological speech impairments. Silent Speech Interfaces (SSIs) emerge as a transformative paradigm that bypasses the acoustic stage by decoding linguistic intent directly from the neuro-muscular-articulatory continuum. This review provides a high-level synthesis of the SSI landscape, transitioning from traditional transducer-centric analysis to a holistic intent-to-execution taxonomy. We systematically evaluate sensing modalities across four critical physiological interception points: neural oscillations, neuromuscular activation, articulatory kinematics (ultrasound/magnetometry), and pervasive active probing via acoustic or radio-frequency sensing. Critically, we analyze the current paradigm shift from heuristic signal processing to Latent Semantic Alignment. In this new era, Large Language Models (LLMs) and deep generative architectures serve as high-level linguistic priors to resolve the ``informational sparsity'' and non-stationarity of biosignals. By mapping fragmented physiological gestures into structured semantic latent spaces, modern SSI frameworks have, for the first time, approached the Word Error Rate usability threshold required for real-world deployment. We further examine the transition of SSIs from bulky laboratory instrumentation to ``invisible interfaces'' integrated into commodity-grade wearables, such as earables and smart glasses. Finally, we outline a strategic roadmap addressing the ``user-dependency paradox'' through self-supervised foundation models and define the ethical boundaries of ``neuro-security'' to protect cognitive liberty in an increasingly interfaced world.

Silent Speech Interfaces in the Era of Large Language Models: A Comprehensive Taxonomy and Systematic Review

Abstract

Human-computer interaction has traditionally relied on the acoustic channel, a dependency that introduces systemic vulnerabilities to environmental noise, privacy constraints, and physiological speech impairments. Silent Speech Interfaces (SSIs) emerge as a transformative paradigm that bypasses the acoustic stage by decoding linguistic intent directly from the neuro-muscular-articulatory continuum. This review provides a high-level synthesis of the SSI landscape, transitioning from traditional transducer-centric analysis to a holistic intent-to-execution taxonomy. We systematically evaluate sensing modalities across four critical physiological interception points: neural oscillations, neuromuscular activation, articulatory kinematics (ultrasound/magnetometry), and pervasive active probing via acoustic or radio-frequency sensing. Critically, we analyze the current paradigm shift from heuristic signal processing to Latent Semantic Alignment. In this new era, Large Language Models (LLMs) and deep generative architectures serve as high-level linguistic priors to resolve the ``informational sparsity'' and non-stationarity of biosignals. By mapping fragmented physiological gestures into structured semantic latent spaces, modern SSI frameworks have, for the first time, approached the Word Error Rate usability threshold required for real-world deployment. We further examine the transition of SSIs from bulky laboratory instrumentation to ``invisible interfaces'' integrated into commodity-grade wearables, such as earables and smart glasses. Finally, we outline a strategic roadmap addressing the ``user-dependency paradox'' through self-supervised foundation models and define the ethical boundaries of ``neuro-security'' to protect cognitive liberty in an increasingly interfaced world.
Paper Structure (64 sections, 2 equations, 4 figures, 5 tables)

This paper contains 64 sections, 2 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Sensing modal Taxonomy of Silent Speech Interface
  3. Neuro-Physiological Interception
  4. Cortical Intent
  5. Neuromuscular Execution
  6. Articulatory Kinematics
  7. The Magnetic Frontier
  8. Anatomical Synergy
  9. Intra-Oral Sensing
  10. Imaging and Optical Sensing
  11. Ultrasound Tongue Imaging (UTI)
  12. Real-time MRI (rtMRI)
  13. Video Lip Reading
  14. Acoustic and Radio Sensing
  15. Active Sonar and RF Micro-Doppler
  16. ...and 49 more sections

Figures (4)

  • Figure 1: Total number of publications in the field of Silent Speech Interface (SSI). The statistics were retrieved from the Web of Science using the themes "Silent speech", "Silent speech interface", "SSI", "Silent speech recognition", "Subvocal speech", "Mute speech recognition", "Non-audible speech", "Articulatory speech recognition", "Articulatory-to-text", and "Speech without sound", covering the period from 2011 to 2025.
  • Figure 2: Framework of this survey paper.
  • Figure 3: The The Neuro-Muscular-Articulatory (NMA) Continuum and SSI Interception Boundaries. The schematic delineates the physiological stages of speech production, from cortical intent and neural signaling to muscular activation and anatomical deformation. The SSI Interception Zone (highlighted in red) defines the functional domain of technologies that bypass the acoustic stage (Traditional ASR) to restore or augment communicative agency via non-vocal biosignals.
  • Figure 4: Overview of Silent Speech Interface (SSI): Key sensing modalities, typical application domains, open challenges and future directions. This taxonomy bridges the gap between raw physiological signal interception and high-bandwidth communicative utility across clinical and ubiquitous computing contexts.