Double Major: Information & Communication Engineering + Computer Science
Passionate about speech data processing, real-time inference, and AI in language learning.
Currently developing DictionDuo – a real-time Korean pronunciation correction AI.
- Exploring speech features like Mel Spectrograms, phonemes, and acoustic patterns.
- Interested in Conformer-based models and end-to-end speech feedback systems.
- Building serverless AI pipelines using AWS Lambda, API Gateway, and SageMaker.
- DictionDuo: Korean pronunciation correction AI powered by Conformer
- Speech error detection & phoneme-level feedback
- Cloud-based inference system with real-time response