Digital Signal Processing Reference
In-Depth Information
4.4
Proto-type Navigation Dialogue
Finally, we have developed a prototype dialog system for data collection
in the car environment [46]. The dialog system is based on the DARPA
Galaxy Communicator architecture [47,49] with base system components
derived from the CU Communicator system [1,17]. Users interacting with the
dialog system can enter their origin and destination address by voice.
Currently, 1107 street names for Boulder, Colorado area are modeled. The
dialog system automatically retrieves the driving instructions from the
internet using an online WWW route direction provider. Once downloaded,
the driving directions are queried locally from an SQL database. During
interaction, users mark their location on the route by providing spoken
odometer readings. Odometer readings are needed since GPS information has
not yet been integrated into the prototype dialog system. Given the odometer
reading of the vehicle as an estimate of position, route information such as
turn descriptions, distances, and summaries can be queried during travel (e.g.,
“What's my next turn”, “How far is it”, etc.).
The system uses the University of Colorado SONIC [23,25,48] speech
recognizer along with the Phoenix Parser[1] for speech recognition and
semantic parsing. The dialog manager is mixed-initiative and event driven
[1,17]. For route guidance, the natural language generator formats the driving
instructions before presentation to the user by the text-to-speech (TTS) server.
For example, the direction, “Park Ave W. becomes 22nd St.” is reformatted
to, “Park Avenue West becomes Twenty Second Street”. Here, knowledge of
the task-domain can be used to significantly improve the quality of the output
text. The TTS system is based on variable-unit concatenation of synthesis
units. While words and phrases are typically concatenated to produce natural
sounding speech, the system can back off to smaller units such as phonemes
to produce unseen words.
5.
DISCUSSION
In this study, we have considered the problem of formulating an in-vehicle
speech dialogue system for route navigation and planning. We discussed a
flow diagram for our proposed system, CU-Move, and presented results from
several sub-tasks including development of our microphone array CSA-BF
processing scheme, environmental sniffing, speech enhancement processing,
