audio segmentation to train Speech Activity Detection (SAD) used in NIST Speech Analytic Technologies evaluation series (OpenSAT)
- Annotated speech events follow Audacity label format: http://manual.audacityteam.org/man/label_tracks.html
- Audio source: http://graphics8.nytimes.com/packages/html/nyregion/20050812_WTC_GRAPHIC/met_WTC_histories_full_05.html