DeafAssist
Index
Privacy & website notes

DeafAssist

Environmental sounds,
shown near the eye.

AI eyewear for Deaf and hard-of-hearing people. It recognises environmental sounds and presents their category and coarse direction near the eye.

Scroll to see the device

A cue, close
to the eye.

A single clear pane presents the cue near the eye. Its compact support connects directly to the frame.

Sound becomes
a visual cue.

DeafAssist concept assembly with a compact housing and single clear near-eye pane

The AI identifies environmental sounds and estimates their coarse direction. The wearer sees a brief cue.

Examine the system

Inside DeafAssist

  1. 01 Stereo sound input
  2. 02 Embedded AI
  3. 03 Near-eye display

An embedded AI model recognises eight sound categories and turns them into a visual cue with coarse direction. A Bluetooth phone companion adds haptic alerts.

Enhanced view of the physical DeafAssist ear-side prototype
Enhanced view of the physical prototype

From a model
to a physical form.

DeafAssist was an IRIS National Science Fair finalist project by Aarush Singh Minhas and Rudra Puri. The work developed into a physical ear-side prototype.

See the development record

Recognition,
at a glance.

88.34%accuracy

What should the
wearer notice?

Future work will focus on comfort, legibility and giving the wearer control over visual attention.