T10
A primer on building wearable vibrotactile devices for improved speech in noise perception
Recent research has demonstrated the efficacy of vibrotactile feedback in enhancing speech perception in noisy environments, benefiting both individuals with normal hearing and those with hearing impairments (Guilleminot & Reichenbach, 2022, doi:10.1073/pnas.2117000119; Rautu et al., 2023, doi:10.1038/s41598-023-43644-3). This approach shows particular promise for populations with age-related hearing loss and cochlear implant users, where additional sensory input can complement auditory processing. However, current vibrotactile solutions often rely on bulky motors or expensive commercial devices that require users to hold the haptic device,, limiting their practical application and accessibility. This approach also misses the opportunity for multi-tactor haptic configurations, which have been demonstrated to improve information throughput in other haptic applications (Tan et al., 2020, doi:10.1109/JPROC.2020.2992561)
We hope to lower the barrier-to-entry of additional exploration in this field by presenting cost-effective approaches to developing multi-channel haptic devices specifically designed for speech-in-noise applications. We outline methods for constructing affordable, wearable solutions that can be readily implemented by researchers in the speech perception field. We share our design and also provide some options for haptics that use commercial-off-the-shelf parts that can be purchased at large online retailers. Our suggested implementation leverages existing audio signals to drive the haptic feedback, eliminating the need for complex signal processing chains or additional sensors. The system architecture employs a multichannel design that enables straightforward synchronization across tactors, crucial for maintaining temporal relationships in speech signals and spatial cues. We achieve this through direct audio-to-haptic conversion with latency below 10ms, ensuring that vibrotactile feedback remains perceptually aligned with acoustic input. This approach allows researchers to readily integrate the system with existing audio processing pipelines while maintaining the temporal precision necessary for speech enhancement applications. Through a systematic review of existing haptic design principles and best practices, we share guidelines for optimizing vibrotactile feedback in speech enhancement applications (Orzech et al. 2024, doi:10.1177/107118132412759).
Additionally, we present the results of a user survey exploring preferences for wearable haptic devices in hearing augmentation, specifically comparing head-worn and wrist-worn configurations designed for speech-in-noise perception and sound localization. Our study includes participants across different hearing profiles, from those with mild hearing loss to users of hearing aids and cochlear implants. We examine key factors including user comfort, signal perception, and integration compatibility with existing assistive hearing devices. Our results provide insight into user preferences and design considerations that may inform future development of haptic solutions for auditory rehabilitation.