This project aims to build an ASR pipeline for European Language (German) and it must be built as a Singularity container on the Case HPC and put into production to process daily incoming files. The project proposes to use different state-of-the-art Speech to Text open source toolkits (Kaldi, DeepSpeech).
It can be divided into following sub-tasks:
- Data cleaning and Feature Extraction
- Language Modeling
- Building Phoneme Dictionary
- Acoustic Modeling