First page Back Continue Last page Overview Graphics

Individuals : Reality Mining


Notes:

For the nodes – I run the audio through a speech detector using a multi-band center clipper– essentially a bank of filters over the speech bandwidth. I preprocess the audio for speech recognition engine and then put the resultant transcriptions into a topic spotter – using the OpenMind CommonSense Data Base – or just bag-of-words. Location information can taking for Access Point information within the wireless network.