piedpiper-2024/verbifyai

AI speech therapist built for kids with special needs powered by Gemini and African Whisper.. # VERBIFY AI Children with speech disorders, such as those diagnosed with autism spectrum disorder (ASD) or dyspraxia, encounter significant challenges in effectively communicating due to difficulties in articulation and pronunciation. Traditional speech therapy methods often fall short, as they may not be readily accessible or tailored to meet the unique needs of each individual. Moreover, consistent one-on-one therapy sessions can be both time-consuming and financially burdensome, limiting accessibility for many children who could benefit from such support. As a result, there is a pressing need for innovative and accessible solutions that address the specific communication challenges faced by children with speech disorders, ensuring they receive the personalized support required to thrive and communicate confidently in their daily lives. Verbify AI presents a solution to the challenges faced by children with speech disorders. By integrating Gemini's advanced AI capabilities into our platform, Verbify AI offers personalized and accessible speech therapy solutions tailored to the unique needs of each child. It is a sound-to-sound AI model utilizing Gemini Pro specifically designed for speech therapy to aid children in improving their speech. This AI transcribes sound into text through an API, allowing for accurate personalization to meet the needs of each child. Trained on various datasets, the model tailors interventions based on the child's current level, providing encouragement on failures and applause upon reaching milestones. Through its empathetic training approach, Verbify AI supports children as they progress through therapy, celebrating achievements and providing personalized feedback. This empowers children with speech disorders to enhance their communication skills, fostering confidence and independence in their interactions. ### simple usage instructions. - `git clone github.com` - run `make build` to install dependenc …