Impressed by his grandparents’ struggles with fashionable expertise as a consequence of motor tremors, highschool junior Dhanvin Ganeshkumar is growing an assistive app known as Swype AI. The appliance goals to make computer systems extra accessible for individuals with motor disabilities, equivalent to these attributable to Parkinson’s illness or ALS. Swype AI will enable a consumer to regulate their pc via a free smartphone app, utilizing easy gestures or voice instructions, bypassing the necessity for a standard mouse or keyboard which may be troublesome for some to function.
Ganeshkumar is growing the app as a for-profit startup, a choice he made after realizing a non-profit mannequin can be more durable to scale. Nevertheless, his major mission is to make sure the expertise is broadly obtainable. The plan is to supply the core performance of Swype AI without spending a dime, making it accessible to underprivileged communities that can’t afford costly assistive units, which might typically price over $1,000. The corporate will generate income via non-compulsory in-app purchases for superior customization options.
Constructed utilizing Python and superior hand-tracking expertise, the mission confronted a major technical hurdle throughout growth. Early variations of the app had a noticeable lag, and the unsteady hand actions of customers with tremors would trigger the on-screen cursor to be erratic and troublesome to regulate. To unravel this, the staff built-in a complicated smoothing filter, a mathematical instrument that intelligently predicts the consumer’s supposed movement and filters out the “noise” from involuntary shakes. By fine-tuning this technique, they have been capable of create a extremely responsive and correct cursor that strikes easily in real-time, offering a dependable and frustration-free consumer expertise.
At present in beta testing, Swype AI has already gained vital traction. Ganeshkumar has consulted with over 15 accessibility organizations to refine the app and has secured round $7,500 in funding and awards to cowl preliminary manufacturing prices. He’s utilizing coding expertise discovered at his STEM-focused highschool and analysis expertise to construct the app, with plans for a public launch inside the subsequent few months. He hopes his journey encourages different younger individuals to develop and share their revolutionary concepts.
Watch the video beneath to see some fast demos of how Swype AI works and head to their web site to study extra about Swype AI.
Supply (1): ARL Now, Supply (2): DevPost