An automatic child-directed speech detector
OpenEdge ABL
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
config
models
scripts
src/python
.gitignore
LICENSE
README.md
requirements.txt

README.md

Automatic Child-Directed Speech Detector

Installation

You have to install Python 2.7 and pip to run the detector.

Then run the following command to install the required python packages.

# Install requirements
pip install -r requirements.txt

You also need to install sox and openSMILE to run the pipeline.

Running the system

To detect child directed speech in an audio file, run the following commands:

cd scripts
python pipeline.py PATH_TO_AUDIO_FILE.wav

Note that the audio files has to be a mono 16-bit wave file with 16khz sample rate. You can use tools like sox to convert the audio, if necessary.

The pipeline writes the annotations to the file AUDION_NAME_cds.txt.

Citing

If you use our tool for research purposes, then we ask that you cite the following paper:

@inproceedings{cds2014,
 author = {Schuster, Sebastian and Pancoast, Stephanie and Ganjoo, Milind and Frank, Michael C and Jurafsky, Dan},
 title = {Speaker-Independent Detection of Child-Directed Speech},
 booktitle = {Proceedings of the IEEE Workshop on Spoken Language Technology},
 year = {2014}
}