Resampling Finished

I finished resampling the audio for the three age ranges. It took a lot longer than I expected. I ended up using the tic and toc functions in MATLAB to measure time:

Resampling Time 0-12: 6.5 hrs
Resampling Time 12-24: 9.5 hrs
Resampling Time 24-36: 11.97 hrs

CHILDES data

The benefit of the CHILDES data is that it’s free and public. The downside is that it’s not very good quality audio. Which again, isn’t a very start to training a speech recognition system. On the other hand, it’s a opportunity to learn some signal processing techniques. For example, silence detection and denoising. However, I will continue with the audio segments as they are for the LTASS experiment. When that’s finished, I can return to trying to improve the quality.

LENA data

Recently I found out it might be possible to acquire some higher quality audio data from the LENA foundation. Though, this data would be of atypical children (i.e., children with autism). So, analyzing that data could be a research project on it’s own.

Best,
EO