Agenda for the 1/19 RT Teleconference

- Attendees:

  - NIST: Fiscus, Rose
  - ICSI: Chuck, Mathew
  - IBM: Makis, Etienne
  - LDC: Glenn
  - UKA: McDonough, Rochet
  - SRI: Stolke
  - CMU: Burger

- Old Issues

  - SASTT status (NIST)
    - JF reported the code development is proceeding.   We expect
      a release by Feb. 2.

  - Standard segmentation for the STT IHM condition
    - During the last telecon, there was general consensus that a
      having a single site provide a "standard" segmentation output
      or a "standard" crosstalk-removed version of
      the test signals would suffice.
    
    - Andreas and Chuck discussed the possibilility or releasing their
      segmentation which they have already done.  See the email list or
      the data website.

- New Issues
  - RT 07 data site operational
    URL: http://www.nist.gov/speech/tests/rt/rt2007/data
    Username: rt07
    Password: It was given to attendees.  It is the same password as the 
              rt06s data website.  Contact Jon F. to get it..

  - Data issues:
    - Suzi will send a speaker ID/audio file lookup table for the CHIL 
      dev data.  (IT is also on the data web site:

  - Conference Room evaluation corpus
    - The corpus will be 3 hours selected from AMI, CMU, NIST and
      VT data
  
  - Do we need another telecon?
    - General consensus to hold one after the SASTT eval tool has been
      released.

  - Release of new NIST meetings for training:
    - Data will be shipped Monday or Tuesday
    - Send Jon F. an email request for the distribution.

- Uncompleted Action Items

  - NIST publishes the Forced Aligned reference files from RT-05 and
    RT-06
    - NIST release the conference meeting data forced alignments on the
      data web page.  

  - CHIL will report on the soon-to-be-released DEV data and it will be
    documented on a web site
    - previously done.

  - Adam J. to circulate suggested AMI Dev/Train divide