AI4KSL — Maseno University's AI Bridge Between Spoken English and Kenyan Sign Language for Deaf Learners

Maseno University's AI4KSL project set out to build assistive AI technology that translates spoken English into Kenyan Sign Language (KSL), rendered through virtual signing avatars, to bridge communication between Deaf students and their hearing peers and teachers in mainstream Kenyan schools.

Over the two-year project (2023–2024), researchers collected sign-language elicitation data — reading and singing tasks, spoken English, and video-recorded KSL words and sentences with accompanying gloss — from 400 Deaf learners and 48 teachers and tutors of the Deaf specialising in sign language, drawn from Kenyan learning institutions. The resulting corpus was released as an open-access dataset and documented in a peer-reviewed paper, with the underlying data validated through stakeholder and town-hall dissemination workshops involving the Deaf community itself.

The AI4KSL translation prototype is intended to be tested with graduate teacher trainees specialising in sign language and with Deaf students in Kenyan schools, with the stated aims of breaking communication barriers, increasing learning outputs and improving bilingual (English–KSL) proficiency.

As a foundational research and dataset project rather than a deployed classroom tool, AI4KSL has not yet published measured effects on Deaf learners' academic outcomes or on teachers' sign-language proficiency; its evidenced contribution to date is the open KSL dataset itself and the validated data-collection process behind it.

Read the full analysis: https://www.maseno.ac.ke/ai4ksl-bridging-language-barrier-using-artificial-intelligence-kenyan-sign-language-among-deaf