Some time ago, I published a paper that described and validated a method for detecting face-to-face communication by processing audio signals from wireless mikes.  See

Corman, S. R., & Scott, C. R. (1994). A synchronous digital signal processing method for detecting face-to-face organizational communication behavior. Social Networks, 16, 163-179.

At the time, the method wasn't really practical, but lately digital recorders that run for long periods of time have become available and cheap.  We've been collecting data in a small organization for about 6 months now using such a system.  We're looking at the perceived vs. observable communication problem and are working on a model dealing with that.

Best regards,

SC

________________________________________________
Steven R. (Steve) Corman
Professor, Hugh Downs School of Human Communication, Arizona State University
http://www.public.asu.edu/~corman
Visiting Professor, Fakultaet fuer Informatik, University of Karlsruhe (Germany)
Vice-chair, Organizational Communication Division, International Communication Association