An immersive experience using AI in tarteel ai app
summary
Tarteel.ai uses supervised machine learning trained on a custom dataset of volunteer Quran recitations with manual annotation to detect pronunciation and Tajweed errors, then scores and suggests corrections in-app. The pipeline uses an acoustic model, Google Cloud speech/labeling APIs and Amazon SageMaker for training, plus optimized Nvidia inference components to run automatic feedback on the phone. The system handles common qira'at variations and outputs phrase-level correction and confidence scores.
AI-written summary · about
chapters (10)
- 0:00Introduction and thanks
- 1:42Tarteel AI solution overview
- 5:11Model training with images
- 8:25Quran recitation accuracy
- 11:09Ramadan project and automation
- 15:12Mood and annotation challenges
- 17:00Algorithm and AI models
- 20:56Inference speed and deployment
- 22:39Future vision and support
- 25:57Quran recitation rules and tajweed
related talks
14mAI for a Billion Voices: Taking AI where there is Maximum Impact
27mIndian Sign Language Recognition(ISLAR)
02h 03mBuilding on AI - Delhi meet-up
16mThe End of Awkward AI Transcriptions
35mAI and Advanced Analytics applied at
25mHigh Impact AI Projects in the Egyptian Government
39mIt's All in the Data: The ML Behind Alexa's AI Systems
44mTransforming education with AI