Page 186 - AI for Good-Innovate for Impact Final Report 2024

AI for Good-Innovate for Impact

Multi-modal content analysis (visual , audio & text) using computer vision and AI dubbing
technology, sign language model training, Public Health (WHO) policy triggers. If given
scholarship we would set up a data simulator which reflects the data generation from a variety of
sources such as audio, visual and text, including historical records, online existing sign language
imagery/ videography in African National sign languages. This would include spatial coverage
of public health institutions and new content alerts from telecom providers via the AI platform.

Prediction on policy triggers in Health and impacts on resource allocation in African Sign
Languages , including readiness will be demonstrated and published with the simulated data
patterns via Public Platforms which can be accessed remotely using multiple Tecomms Service
providers with last mile connectivity.

42�3� Use case requirements

• REQ-01: Live Sign Language Translation Tool - Develop an AI-powered live sign language
translation tool capable of translating between at least 25 African sign languages and
spoken/written language in real-time.
• REQ-02: The application of the use case: Purple Signs - ITU Telecom World.
• REQ-03: Content Creation and Dubbing - Use AI to create and dub content into sign
language, making information accessible to the deaf community.

42�4� Sequence Diagram

42�5� References

[1] HIPH and Purple Signs Global

[2] Pharmaceutical Society Of Zimbabwe Joint Congress [PSZ] , 2018 (PSZ , 2019)

[3] Zimbabwe Medical Association Annual Scientific Congress Journal , 2019

170