README

Authors

Alexandre Nanchen, Christine Marcel

Description

This is the README for the Automatic Speech Recognition Tools.

This project contains various scripts in order to facilitate the preparation of ASR related tasks.

Current tasks are:

1. Sentences extraction from pdf files
2. Sentences classification by language
3. Sentences filtering and cleaning

Document sentences can be extracted into single document or batch mode.

For an example on how to extract sentences in batch mode, please have a look at the run_data_preparation_task.sh script located in examples/bash directory.

For an example on how to extract sentences in single document mode, please have a look at the run_data_preparation.sh script located in examples/bash directory.

There is also an API to be used in python code. It is located into the common package and is called DataPreparationAPI.py

Docker

To build a docker image of asrt:

docker build -t asrt .

To run run_data_preparation.py using docker image, where your test file is docker-example/research.txt:

docker run -d -v $PWD/docker-example:/usr/local/asrt/data asrt -i data/research.txt -o /usr/local/asrt/data
# Or with the same pdf as above example:
docker run -d -v $PWD/docker-example:/usr/local/asrt/data asrt -i data/Research.pdf -o /usr/local/asrt/data

The output will then be in docker-example/sentences_*.txt