Hackathon: Silent Speech Interfaces

​SSI's decode whispered speech to text, making them a promising human computer interface.

​Main advantages:

• ​Vocal cords are not vibrating when whispering, allowing for effortless voice input over long hours.

• ​A person sitting directly next to you is not able to hear what you're saying when using a SSI, allowing for private use in public spaces.

​Groups hack on either:

• ​Hardware: CAD design & mechanical principles for acoustic amplification of speech (we provide a 3D printer, Arduino, Microphones)

• ​Machine Learning: Fine-tuning voice recognition models on whispered audio.

​Demo: [Epiph.Tech](https://www.linkedin.com/posts/raghav-sanagavarapu-b914481b7_ai-in-your-head-imagine-accessing-the-world-activity-7160343341316919296-RLzC?utm_source=share&utm_medium=member_desktop)

Paper: https://ieeexplore.ieee.org/document/1200069