Arabic-English-Subtitling

Automatic Video Subtitling System for Arabic/English Using Speech Recognition

The goal of this project is to develop an application that allows subtitling of English and Arabic. The application will provide the user with a full subtitle file at the end of the execution. The subtitling will be based on the CMU-Sphinx speech recognition engine which is open source GNU software.

In this project, the ASR engine will be mostly used as a black box and the role of the student would be to understand how to integrate the engine with his application.

Main 6 Steps for subtitling

1. Extract audio from video FFMPRG
2. Convert audio to 16khz one channel FFMPRG
3. Finding out where each speech pause in the audio LIUM
4. Segmenting the audio at each pause using FFMPRG
5. Speech recognize each segment ( CMU SPHINX )
6. Save to an srt file

3rd Libraries used

1. LIUM: Detects segments
2. FFMPEG: Converts and segemetns audio
3. SPHINX: Speech Recognition Engine

Scripts used

1. AudioSegmentor: Cuts audio to segments using LIUM & FFMPRG
2. SubtitleWriter: Write the subtile file

Items needed for speech recognition:

1. The decoder source code
2. The language dictionary
3. The acoustic model
4. The language model
5. The test data

Contact

If you want to contact me you can reach me at [email protected].

License

This project uses the following license: GNU.