Shaed

GestureCue Accessibility – AI-Powered Visual Communication GestureCue Accessibility is a macOS application designed to make online presentations more accessible for Deaf and hard-of-hearing audiences. During online presentations, Deaf participants often need to divide their visual attention between the presenter, sign language, captions, and slides. GestureCue addresses this challenge by making key information easier to follow through clear, AI-generated visual cards. Using Meta AI models, GestureCue transforms ordinary educational content into short, accessible messages and relevant illustrations. Presenters can review, edit, regenerate, and save cards before presenting. During live Zoom presentations, GestureCue uses webcam-based hand gesture recognition to display, switch, and hide visual cards without requiring keyboard interaction or voice commands. Flexible video layouts allow presenters to position cards while preserving visibility of the presenter and signing area. Built with SwiftUI, Swift 6, Apple Vision, and AVFoundation, GestureCue combines generative AI, computer vision, and accessible visual communication. The project is inspired by my experience working with Deaf and hard-of-hearing communities in Sweden. My goal is to make digital learning and communication more inclusive, independent, and accessible for everyone.