The Multimodal Reading Tutor is an interactive and engaging tool to assist elementary students in developing their english language skills using Intel's AI PC / NUC and Intel technologies.
- Gives examples of simple sentences
- Uses speech recognition to hear that the student is reading correctly
- Uses pictures to describe the word when it hears long hesitation / mumbling / mispronunciation
- Makes a picture book (each sentence gets a picture)
- It reads to you or you read to it
- utilizes simple sentence feature
- Asks you questions about the reading to develop analytical thinking skills
- Teaches letters of the alphabet & phonemes
- Listens to user say those letters and their sounds
- Intel® OpenVINO™ Toolkit, Intel oneAPI, PyTorch, HuggingFace datasets
- Meta AI’s Wav2Vec speech & phoneme recognition
- Fine tuned Llama for making reading assignments & question prompts
- DALL-E Mini for picture generation
React frontend, Django backend
accessible and affordable tutoring option for students, particularly those in under-resourced areas.
Encourages interactive and multimodal learning, making reading enjoyable and engaging for young learners.
Offers a scalable solution that can be deployed across various educational institutions to support widespread reading development