← ALL PROJECTS
TreeHacks 2024 · Winner, Most Creative Hack & Human Capital Track

Show and Tell

A wearable glove and mobile app that translates American Sign Language into text and speech in real time, and carries the signer's emotional tone through to the listener.

My RoleHardware + wearable UI
Team5 people
Built WithESP32, Hume AI, IMU, OpenAI
Codegithub.com/raviriley/aidactyl

Overview

A lot of existing sign language technology is built to help hearing people understand the deaf, rather than helping deaf people communicate on their own terms. Show and Tell flips that: a glove fitted with flex sensors and an IMU reads the signer's gestures directly, and a companion video call app converts that into text and speech, including the emotional tone behind it.

What I worked on

  • Wearable hardware. Embedded Arduino electronics and strain gauges into a wearable glove format.
  • User interface. Developed the interface used to interact with the glove's output.
  • Embedded systems. Helped build the link between the sensor data coming off the glove and the rest of the system.

How it works

The glove reads flex and motion data to determine which ASL gloss the wearer is signing, then sends that data to a mobile video call interface. During a call, the recognized signs appear as text and are spoken aloud, with emotional cues from a computer vision model layered into the delivery so the listener gets both the words and the feeling behind them.