Note: Descriptions are shown in the official language in which they were submitted.
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
INTELLIGENT CALL PROCESSING
PLATFORM FOR HOME TELEPHONE SYSTEM
' TECHNICAL FIELD
The present invention relates generally to a centralized control system for a
group of
telephones located, for example, in a home, and more particularly to the use
of voice commands
in conjunction with voice recognition and speaker verification techniques to
provide a variety of
telephone services for the users of such telephones.
BACKGROUND OF THE INVENTION
Conventional telephone systems for the home or office include one or more
telephone
sets normally connected to a telephone switching center (i.e., a central
office, PBX, or key
telephone system) through a cable. With the advent of economical and reliable
digital
technology, recently it has been possible to provide enhanced functionality to
the typical
telephone. Thus, many companies are now manufacturing and marketing such
devices as
videophones, display phones, digital answering machines, screen-based, menu-
driven phones and
the like. These devices facilitate provision of various caller services
including call forwarding,
call waiting, voice store and forward, billing assistance, caller
identification, and other functions.
Although such devices provide significant operational capabilities as compared
to conventional
telephone sets, there are a myriad of service offerings and the various types
of equipment have
yet to be successfully integrated into a single system that can be
conveniently or selectively
controlled by the telephone users.
CA 02206870 2005-O1-18
Thus, there remains a need to provide an intelligent call processing apparatus
for
interfacing to, to be controlled through, and to control a plurality of
telephones located in
a defined area, such as a home, to facilitate enhanced telephone service
offerings.
BRIEF SUMMARY OF THE INVENTION
In accordance with one aspect of the present invention there is provided a
control
platform for use between a remote switching system and an on-site telephone
system
including at least one telephone, the platform comprising: means for
interfacing the
platform directly to the on-site telephone system; a trigger re<;ognizer for
recognizing
voice or DTMF command triggers input through the at least one telephone;
announcing
means for providing predetermined messages; processor means, operative under
the
control of a program stored therein and responsive to receipt of an incoming
call from the
remote switching system, for identifying a source identifier o:f the incoming
call; and the
processor means, further operative under the control of the program and
responsive to
receipt by the trigger recognizer of a first voice or DTMF command trigger
during the
ringing stage of the incoming call, for controlling the announcing means to
announce to a
user an identification associated with the source identifier of the incoming
call.
In accordance with another aspect of the present invention there is provided a
centralized control platform for providing a variety of telephone services to
a local
telephone system having at least one telephone, comprising: a local interface
for
connecting the platform to the at least one telephone of the local telephone
system; a
communications circuit for interfacing the platform to external communications
means;
and processing means interconnected with the Local interface and
communications circuit
responsive to receipt of an incoming call from the external communications
means for
providing a user with command triggered access to a plurality of telephone
services, the
processor further comprising: trigger recognition means for recognizing voice
or DTMF
command triggers input by the user through the at least one telephone during
the ringing
2
n~" .~ , M ~ ,.,w, ~. na,.wc .. ,"~a,ur, " , .. _ . ....w, r . ,.....~..w".
>.:.s..,»~.,. . .,...».o..a ,
,tw,~,..a~~~sr.,ira_..,~mr~riE cwwx»-,~,anaww:~.~,wuee,.~,.....~-www~.s,~z;.o--
:..u.s,:»..w~m~.~~.,~.~,nm.;~c_,a~..mrrm..xa.:.u;~r~xmn
CA 02206870 2005-O1-18
stage of the incoming call; and announcing means for providing predetermined
messages
to the user.
BRIEF DESCRIPTION OF THE DRAWINGS
For a more complete understanding of the present invention and the advantages
thereof, reference should be made to the following Detailed Description taken
in
connection with the accompanying drawings in which:
FIGURE 1 is a block diagram of a telephone system incorporating the basic
version of the intelligent call processing platform of the present invention;
FIGURE 2 is a block diagram of a telephone system incorporating the enhanced
version of the intelligent call processing platform of FIGURE l;
FIGURE 3 is a detailed system diagram showing the components of the
intelligent
call processing platform of FIGURE 1 and FIGURE 2;
3
. . wer , . ..;.va.a~:1 ,.,sa'~2w, .x3k, _, p~...,c: W=~.-.e
i.>k:,Sk'~~k.:.'~u.,n msa.. a..-_.~ ~ ....... "." w ~..~..~.~...x.~w
g~c._yvce~,py,~~t.~~ _ aVU _ .. ap : mesz..
CA 02206870 1997-06-03
WO 96/18258 PCT/US95l16391
FIGURE 4 is a flow chart of the executive routine from which all other
routines are
accessed;
FIGURE 5 is a flow chart of the incoming call handler routine for directing
and/or '
controlling incoming telephone calls;
FIGURE ~ is a flow chart of the routine for playing messages to incoming
calls;
FIGURE 7 is a flow chart of the routine for recording messages from incoming
calls;
FIGURE 8 is a flow chart showing the originating call/command handler routine
which is
accessed by the executive routine of FIGURE 4 for defining operations that
occur when the
handset of a connected telephone has been lifted;
FIGURE 9 is a flow chart of the routine for acquiring spoken commands;
FIGURE 10 is a flow chart of the spoken digit collection routine used to
obtain spoken
numeric sequences;
FIGURE 11 is a flow chart showing the routine for providing speed dialing
capabilities
from spoken key words;
FIGURE 12 is a flow chart of the routine for providing speed dialing
capabilities from
speaker-dependent words stored by the user and associated with each telephone
number or
numeric sequence to be dialed;
FIGURE 13 is a flow chart showing the routine for programming speed dial
numbers
using key words;
FIGURE 14 is a flow chart showing the routine for programming speed dial
numbers
using numeric memory locations;
FIGURE 1 S is a flow chart showing the routine for programming speed dial
numbers
using speaker-dependent directory words;
4
CA 02206870 1997-06-03
WO 96!18258 PCTlUS95/16391
FIGURE 16 is a flow chart showing the routine for changing programming
parameters
through both local and remote access;
FIGURE 17 is a flow chart of the routine for playing back messages from
callers; and
FIGURE 18 is a flow chart of the peripheral control routine used to operate
lights,
thermostats, alarm systems, televisions, stereos, or other electronic devices.
Similar reference characters refer to similar parts throughout the several
views of the
drawings.
DETAILED DESCRIPTION
As seen in FIGURE 1, the intelligent call processing platform 2 of the present
invention
is preferably implemented in a home which typically includes one or a
plurality of telephone sets
4a...4n. FIGURE 2 shows an enhanced version of FIGURE 1 and allows for
additional features.
The core system components shown in FIGURE 3 for both FIGURE 1 and FIGURE 2
are the
same with the exception that the system depicted in FIGURE 2 is installed in a
personal
computer or any processor controlled system and utilizes the optional PC
interface 36. Although
the present invention is shown in the environment of a home, it should be
appreciated that the
intelligent call processor is likewise useful whenever it is desired to
provide a number of
telephone sets with certain enhanced features as will be seen. Thus, the
example of an
intelligent home system is merely exemplary.
The intelligent call platform is preferably connected between a switching
system or
switching central office and the telephone sets 4. As seen in FIGURE 1 or
FIGURE 2, the
' platform is connected to all of the telephone sets, although this is not a
requirement of the
system. The method of connection between the platform and telephone sets may
be serial or
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
parallel. The switching office is typically a local telephone operating
company central office
operating within a so-called local access and transport area (LATA) although
this is not the only
configuration in the invention. When long distance service is required, the
call is usually '
transmitted through the operating company central office to a point of
termination in the
originating LATA at which it is picked up by a common earner and passed by
that carrier on to a
termination point in a local or distant LATA. Such operation is conventional
in the art.
Moreover, while the physical connections shown in FIGURE 1 and FIGURE 2 are
typical it is
envisioned that the home be connected to the switching office alternatively
via a coaxial cable,
fiber optic cable, wireless link or other suitable connection. Such
alternative communication
channels are also known.
The intelligent processing platform shown in the system diagram of FIGURE 3
includes a
voice recognition algorithm as well as a speaker verification algorithm to
facilitate
voice-controlled access and control over various service offerings. The
platform may
alternatively include DTMF digit recognition algorithms to provide access and
control over the
service offerings. The platform includes a digital signal processor (DSP) 42
for controlling the
overall operation of the system. The DSP includes data, address, and control
buses represented in
FIGUR~3__ A~ seen- inFIGURE 2,the enhanced -platform- 10 may-also include-
conventional
input/output devices such as a keyboard 12 and display terminal 14 and may
additionally include
a mouse, microphone, speech generator, and printer. A communications interface
52 interfaces
the platform to the internal telephone sets and a communications (CØ)
circuit 48 interfaces the
platform to external telephone lines. If alternative communication links (such
as coaxial cable)
are used, the interface 48 will include additional circuitry to interface the
additional signaling '
(such as video or the like) to the intelligent processing platform.
6
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
The platform also includes a voice and/or video store and forward capability
for recording
messages. Random access memory ("RAM") 32 is connected to the DSP by the
address, data,
and control buses for providing temporary storage of data processed thereby.
Read only memory
" ("ROM") 28 is likewise connected to the digital processor for providing
nonvolatile operational
code. ROM 34 is likewise connected to the digital processor for providing
permanent storage of
speaker-independent recognition tables and prestored messages for instructing
the user on his/her
progress when operating the system. Verification data. is stored in
programmable nonvolatile
FLASH Memory 30. For the enhanced version, disk storage on the personal
computer 10 may
support control programs including the voice recognition algorithm and the
voice verification
algorithm as well as suitable control programs (not shown). A voice command
system of this
type is described in copending application Serial No. 07/523,486, filed May
15, 1990, to Hunt et
al., titled "Simultaneous Speaker-Independent Voice Recognition And
Verification Over A
Telephone Network," assigned to the assignee of the present invention and
incorporated herein
by reference.
The voice recognition algorithm may be based on speaker-dependent or
speaker-independent discrete, speaker-independent phonetic, connected or
continuous
recognition. Such techniques are known in the art. Where reduced cost is
required,
speaker-dependent algorithms are preferably used. A predetermined set of
commands may be
used in the system and/or commands may be created in a known manner. In
particular, where the
user desires to create special commands, the system prompts the user to speak
the command a
number of times while the recognizer is "trained" in a conventional manner. A
template (e.g.,
speech reference data) is created for each command and store in memory. and
the user typically
will then enter the telephone number associated with such command for later
retrieval whenever
7
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
the command is recognized. Thus a user may create a number of special call
commands, e.g.,
"account balance," "directory assistance," "grandma," etc., the recognition of
which control the
processing platform to dial a stored number and/or dialing sequence associated
therewith.
The platform shown in FIGURE 3 also includes a speaker verification reference
database
stored in FLASH Memory 30 comprising speaker verification class reference data
collected from
users authorized to access the service offerings. In this way, the platform
allows control over the
service offerings by selected members of the home. Thus, for example, one or
more family
members may enroll his or her "voice" in the speaker verification reference
database and thus, as
will be seen, exercise control over the ability of other members (such as
children) to access
certain services. Thus, certain of the service offerings may be programmed to
be accessible only
if voice commands are both recognized and/or verified to have come from a
person authorized to
obtain the service. The use of speaker verification thus provides an
additional level of security
which has numerous advantages. For example, parents may program the platform
to prevent
children from making unauthorized calls since such calls will only be allowed
if the voice
command used to institute the call is both recognized and the speaker is
authorized for the
service. Additionally, the speaker verification capability will provide
security for remotely
accessing recorded messages and will allow user specific access to the
recorded messages.
Recorded messages will be stored in the Flash Memory of FIGURE 3 for systems
as shown in
FIGURE 1, but may optionally be stored on the disk drive of the personal
computer 10 of
FIGURE 2. or any other system memory.
In a caller identification application, the database in the platform may be
programmed to
include one or more telephone numbers or groups of telephone numbers that the
homeowner
desires to block; alternatively. the database may be programmed to identify
those telephone
8
CA 02206870 1997-06-03
WO 96118258 PCT/US95/16391
J
numbers from which the caller desires to receive calls. Additionally, specific
outgoing calls may
be blocked respectively for each user. When undesirable calls are received,
the platform
redirects the calls to the play message routine of FIGURE 6 or the record
routine of FIGURE 7 to
play an audio and/or video announcement message and/or to record a message for
the
homeowner. The calls are identified using the Caller ID number transmitted
with each telephone
call.
The platform is preferably accessible for each of the telephones in the home
(or the
platform itself via the microphone) to facilitate voice control of such
functions as dialing, call
forwarding. voice store and forward, smart ring, caller identification, call
waiting, message
playback, and the like. Incoming calls may be directed to specific telephones
or messaging
centers using smart ring or incoming caller identification. Alternatively, the
telephone systems
may be accessed by push button when DTMF recognition procedures are used.
Refernng now to FIGURES 4 through 18, wherein there is illustrated a plurality
of flow
charts illustrating the operating routines enabling voice or DTMF control of a
variety of
telephone services through the Intelligent Call Processor Platform. FIGURE 4
illustrates the
executive routine through which all calls and all command routines are
accessed. The system
monitors at step 102 for an incoming call. When an incoming call is detected,
the call is
forwarded to the incoming call handler routine which will be more fully
discussed with respect to
FIGURE 5. When no incoming calls are detected control passes to inquiry step
104 to determine
if an off hook condition exists. When no off hook condition exists. control
returns to step 102,
and the executive routine continues to monitor for either an incoming call or
telephone off hook
condition. If a telephone off hook condition is detected, control passes to
the originating
call/command handler routine described with respect to FIGURE 8.
9
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/i6391
Refernng now to FIGURE 5, there is illustrated the incoming call handler
routine. The
incoming call handler routine directs and/or controls incoming telephone calls
and assists in
routing the call to the proper telephone service. A test performed at step 202
determines if the
smart-ring service is active for the incoming call. If so, the smart-ring
table stored in FLASH
MEMORY 30 is accessed at step 204 to determine the proper telephone port to
signal for the
incoming call. The telephone port flag corresponding to the determined port is
set at step 206.
Once the telephone port flag is set or inquiry step 202 determines the smart-
ring feature is not
active, inquiry step 208 determines if the Caller LD. function is active for
the incoming call.
When the Caller LD. function is active, the Caller LD. table is checked at
step 210 to determine
the telephone port to signal and the identity of the caller based upon the
Caller LD. transmitted
i
with the call. The corresponding telephone port flag is set at step 212.
After setting the correct telephone port flag or determining the Caller LD.
function is not
active, step 214 determines if a prerecorded message is to be played after the
second or fourth
ring. If such a message is to be played, the platform checks a function table
and answer flag
based upon the Caller LD. and/or the smart-ring flags at step 216 to determine
whether or not to
answer the call. If the call should be answered, test step 218 passes control
to step 220. When
the call is answered, the play and record message flags are checked at step
220, and decision step
222 transfers control to either the play message routine of FIGURE 6 or to the
record incoming
call routine of FIGURE 7 based upon the settings of the control flag. A
platform user may also
set the flags after having an incoming call identified to the user based on
the Caller LD. The
received Caller LD. number may be looked up in the Caller LD. table and the
user notified of the
name associated with the Caller LD., or if the LD. number is not in the table,
the user may be
notified with the calling number.
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
If the call is not answered or if a message is not to be recorded or played
after the second
or fourth ring, control passes to step 224 to scan for an incoming ring
signal. When an incoming
ring signal is detected, the processor checks the telephone port flags at step
226 to determine
which telephones connected to the platform should be rung. Test step 228
determines if a ring
parameter conflict exists, and if such a conflict exists, the smart-ring
parameter takes priority at
step 230. The ring signal is generated at step 232 to the selected
telephone(s). When no ring
parameter conflict exists, the ring signal is also generated at step 232 to
the selected telephone(s).
Next, inquiry step 234 determines if the telephone is in an off hook
condition. When an
off hook condition does not exist, the incoming call ring count is incremented
at step 236 for the
present call. Inquiry step 238 determines if the call answer flag has been
set, and if not, performs
a test at step 240 to determine if the call ring count equals ten. If the call
ring count equals ten,
control passes to the system access routine of FIGURE 16 to enable an outside
caller to remotely
access the system. If the call ring count does not equal ten or if the answer
flag has not been set,
a test is performed at step 244 to determine the existence of an incoming call
or an off hook
condition. Control then passes back to the executive routine of FIGURE 4.
Returning now to inquiry step 234, if the telephone is determined to be in an
off hook
condition, the ring voltage to the telephone is discontinued, the smart-ring
and LD. port flags are
reset and the audio path connected at step 242. The telephone is tested for an
on-hook condition
at step 248 until an on-hook condition is detected. Control then returns to
the executive routine
of FIGURE 3.
Referring now to FIGURE 6, there is illustrated the routine for playing
messages to
incoming calls. Once inquiry step 222 of FIGURE 5 determines a message is to
be played, the
11
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
message in progress flag is set at step 502. The phone is indicated at step
504 to have an
off hook condition at the central office and a prerecorded outgoing message is
played based upon
the Caller LD. and/or ring cadence. A test is performed at step 506 to monitor
for an interrupt
condition of either a DTMF "0" digit or the spoken command "CONTROL
FUNCTIONS." If an
interrupt condition is detected, the outgoing message is interrupted and
control passes to the
system access routine of FIGURE 16. Otherwise, after the prerecorded message
has been played,
the call is disconnected, an on-hook condition is indicated to the central
office and the message
in progress flag is reset at step 508. Control then returns to the executive
routine of FIGURE 4
to monitor for additional incoming calls or telephone off hook conditions.
Referring now to FIGURE 7, there is illustrated the record incoming call
routine. This
routine records an incoming call routed to its control functions. If inquiry
step 222 of FIGURE 5
indicates a call is to be recorded, the recording in progress flag is set at
step 602. Next, the
platform notifies the central office of an off hook condition, plays a
prerecorded outgoing
message based upon the Caller LD. and/or ring cadence and records a message
from the
incoming call at step 604. Inquiry step 606 monitors for an interrupt
condition consisting either
of a DTMF of "0" digit or the spoken command "CONTROL FUNCTIONS." If one of
the
interrupt conditions are detected, control passes to the system access routine
of FIGURE 16.
Otherwise, the recorded message is stored at step 608 based upon the Caller
LD. and/or ring
cadence. The svstem then resets the recording in process flag, disconnects the
call and provides
an on-hook indication to the central office at step 610.
Refernng now to FIGURE 8, there is illustrated the originating call/command
handler
routine. This routine enables access of telephone functions from within the
platform once the '
handset of a telephone connected to the platform has been placed in an off
hook condition. Once
12
CA 02206870 1997-06-03
WO 96/18258 PCT/US95116391
an off hook condition is detected by inquiry step 104 of FIGURE 4, a dial tone
is generated at
step 304 to the user. This may be the dial tone from the central office or an
artificially generated
dial tone. Inquiry step 308 tests for receipt of a DTMF digit. If a DTMF digit
is detected, the
dial tone is discontinued, and the provided DTMF digits are collected at step
316. Next, a check
is performed at step 320 for any incoming calls. If no incoming calls are
detected, tests are
performed at step 322 and 324 to determine if the message in progress or
recording in progress
flags are set.
If neither of these flags are set, the outgoing line is seized, and the
collected DTMF digits
are outpulsed at step 332. If an incoming call is detected, or the message in
progress or recording
in progress flags are set, a message is played at step 326 asking the user to
hold. Test Step 328
monitors for disconnection of the incoming call, and seizes the outgoing line
at step 332 once the
incoming call is disconnected. Once the outgoing line is seized, the DTMF
digits are outpulsed
and the audio path is connected. Step 334 monitors for an on-hook condition
and passes control
back to the executive routine of FIGURE 4 when an on-hook condition is
detected.
If no DTMF digits are detected at step 308, inquiry step 310 tests for a
speech command
on the telephone line. When a speech command is detected. the dial tone is
discontinued at step
314 and control passes to the spoken command control routine of FIGURE 9. When
no speech
commands are detected at step 310, inquiry step 312 monitors for an on-hook
condition. If no
on-hook condition exists, control returns to step 308 and the platform
continues to monitor for
DTMF digits or voice commands. When an on-hook condition is detected, the
routine returns to
the executive routine of FIGURE 4.
Referring now to FIGURE 9, there is illustrated the spoken command control
routine.
This routine enables a user to vocally select a variety of telephone services.
The spoken
13
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
command detected at step 310 of FIGURE 8 is retrieved at step 802. A variety
of tests are
performed at steps 804, 806, 808 and 810 to determine which spoken command has
been
received. If a "DIAL" command is detected at step 804, control passes to the
spoken digit
collection routine of FIGURE 10. If a "CALL" command is detected at step 806,
control passes
to the key word speed dialing routine of FIGURE 11. The detection of a
"DIRECTORY"
command at step 808 passes control to the speed dialing from user director
routine of FIGI1RE
12. Finally, if a "SELECT" command is detected at step 810, the command is
processed within
the local branch of the system access routine of FIGURE 16. If none of these
voice commands
are detected, an error condition exists and the "Error, please repeat" message
is played to the
caller or user at step 812. Control then returns to step 802 to monitor for
the repeated spoken
command.
Referring now to FIGURE 10, there is illustrated the spoken digit collection
routine.
This routine accepts a spoken digit string, representing a telephone number,
from a user and
forwards the number for dialing. A user is prompted at step 1002 with a "Phone
number please"
message to begin collection of the spoken number string. Inquiry step 1004
determines if a
spoken digit has been collected. If a digit has not been collected. inquiry
step 1006 determines if
a time-out period for collection of the digit has elapsed. Once the required
time-out period
elapses a "Please enter digit" message plays at step 108 to prompt the user to
enter or re-enter the
required digit. Once step 104 determines a digit has been collected, a short
beep is generated to
the user at step 1010 to indicate the digit has been received and recognized.
Inquiry step 1012
calculates if the previously collected spoken digit was the last digit
expected in the string. If
not, the digit is saved in a string buffer at step 1014 and control passes
back to step 1004 to
accept the next digit in the string.
14
CA 02206870 1997-06-03
WO 96/18258 PCT/US95I16391
Once the last digit of the numeric string has been collected, the speaker-
independent
recognizer is engaged at step 1016. Step 1018 monitors for the spoken command
"VERIFY." If
the "VERIFY" command is not received, a message instructing the user to repeat
the "VERIFY"
command is played at step 1020. Once a "VERIFY" command is recognized, the
platform
repeats the digit string to the user at step 1022.
Inquiry step 1024 monitors for the "STORE" voice command and passes control to
the
key word speed dial number storage routine of FIGURE 13 when the command is
detected. If no
"STORE" voice command is detected, step 1026 monitors for a "SEND" voice
command. Once
a "SEND" voice command is detected, control passes to step 320 of the
originating
call/command handler routine of Figure 8 to output the collected digit string.
Otherwise, the
message "Error, please repeat" prompts the user at step 1028 to provide the
"STORE" or
"SEND" command.
Referring now to FIGURE 11, there is illustrated the key word speed dialing
routine.
This routine enables a user to access a previously stored telephone number in
response to a
spoken speaker-independent key word. At step 1102. the platform plays a
message "Calling,"
prompting the user to begin the key word dialing sequence. The speaker-
independent recognizer
is engaged at step 1104 to obtain a user response. The speaker-independent
recognizer monitors
at step 1106 for the spoken command "MEMORY." If the "MEMORY" command is not
detected, the user must speak a valid key word at step 1108 to speed dial the
associated number.
Examples of valid key words include "BANK", "ACCOUNT BALANCE", "OFFICE",
"TIME",
"SECRETARY", "FRIEND", "WOR.K" AND "INFORMATION." This list is not intended to
be exhaustive and any number of key words may be used. Inquiry step 1110
determines if a
valid key word has been spoken. If the platform, does not recognize the key
word, step 112
CA 02206870 1997-06-03
WO 96/18258 - PCT/US95/16391
prompts the user with the message "Error, please repeat." When the key word is
accepted as
valid, a "Calling" message is played at step 114.
If the command "MEMORY" is detected at step 1106, the user is prompted with
message
"Which memory" and the speaker- independent recognizer is engaged. At inquiry
step 1130 the
speaker-independent recognizer determines if a two-digit memory location
number has been
spoken by the user. If not, the platform performs a test check at step 1132 to
determine if the
time out period has elapsed. Once the time out period elapses, the user is
prompted at step 1134
with a "Please enter next digit'' message. Once the 2-digit memory location
number is collected,
the telephone number associated with the memory location number is recalled at
step 1136 from
FLASH memory 30. The user is then notified at step 114 with a "Calling"
message.
Once a key word or memory call location is collected, the speaker-independent
voice
recognizer is engaged at step 1116 The speaker-independent voice recognizer
monitors at
inquiry step 1118 for a spoken command "VERIFY", and when the "VERIFY" command
is
detected, the stored number associated with the key word or memory location is
repeated to the
user at step 1122. At inquiry step 1124, the speaker-independent recognizer
monitors for a
spoken command "CLEAR" indicating that the user desires to clear the number
retrieved and
start again. If so, the taped message "Ready" is played at step 1126 and
Control returns to step
1108. If neither the "CLEAR" command nor the "VERIFY" command are recognized,
inquiry
step 1120 determines if the voice command "SEND" has been spoken. If not, the
user is
prompted at step 1138 with an "Error, please repeat'' message and control
passes back to step
1116. Otherwise, control passes to step 320 of the originating call/command
handler routine of
FIGURE 8 to call the selected number.
16
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
Referring now to FIGURE 12, there is illustrated the routine for speed dialing
from a
pregenerated user directory. This routine enables a user to speed dial a
number from a
preexisting directory using speaker-dependent names. The user is prompted at
step 1202 with a
' "Name please" message. Step 1202 also engages the speaker-dependent
recognizer. At inquiry
step 1204, a determination is made if a spoken name has been collected and
recognized by the
speaker-dependent recognizer. If no name has been recognized, step 1220
determines if the
timeout period has elapsed, and if not, prompts the user at step 1222 with an
"Error, please
repeat" message. Once the timeout period elapses, the user is further prompted
with the "Name
please" message of step 1202.
Once the name is collected at step 1204, the telephone number associated with
the name
is recalled at step 1206 from FLASH memory 30. The stored number is played to
the user at step
1208 after which the user is prompted with a "Correct" message. Inquiry step
1210 then
monitors for a spoken command "CORRECT." Upon detection of the "CORRECT"
command,
the user is provided with a "Dialing" message and control passes to step 320
of the originating
call/command handler routine of FIGURE 8. When the "CORRECT" command is not
received,
control returns to step 1202 to have the user repeat the name.
Referring now to FIGURE 13, there is illustrated the key word speed dial
number storage
routine. This routine enables a user to associate a selected speaker-
independent key word with a
telephone number and have the number and associated key word stored in memory.
Initially, the
user is prompted at step 1402 with a "Storing'' message as the digit string
from FIGURE 10 is
saved. and the speaker-independent recognizer is engaged. The speaker-
independent speech
recognizer monitors at step 1404 for the spoken command "MEMORY." This enables
the user
to stop the number at a location identified by a two-digit memory code. If the
"MEMORY''
17
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
command is received, control passes to the memory speed dial number storage
routine of
FIGURE 14. Otherwise, the spoken independent recognizer monitors at step 1406
for a spoken
command "DIRECTORY." If the "DIRECTORY" command is detected, control passes to
the
director speed dial number storage routine of FIGURE 15, and the user may
store a directory _
name or a number associated with a directory name.
When neither a "MEMORY" nor "VOICE" command is detected, the recognizer
monitors at step 1410 for entry of a new speaker-independent key word. When no
new key word
is detected, an "Error, please repeat" message prompts the user to enter a
spoken command or
key word at step 1420 and control returns to inquiry step 1404. When a new key
word is
collected, the user is prompted at step 1412 with a "Storing (key word)
correct" message.
Inquiry step 1414 then determines if the received key word was correct. If
not, a "Location,
please" message prompts the user at step 1418 to repeat the key word and
control returns to step
1410. Once the correct key word is entered, the system advises the user at
step 1416 with a
"Storing" message, and the number and associated key word are stored in FLASH
memory 30.
Control then passes to step 304 of the originating call/command handler
routine of FIGURE 8.
Referring now to FIGURE 14, there is illustrated the memory speed dial number
storage
routine. This routine allows the user to store a particular number in a
selected memory location
having an associated two-digit memory code. The user is prompted at step 1502
with a "Which
memory?" message and the speaker-independent recognizer is engaged. At inquiry
step 1504.
the speaker-independent voice recognizer monitors for receipt and recognition
of a 2-digit
memory location number. If both digits have not been received, the routine
continues to cycle
through step 1506 until a timeout period has elapsed. If both digits have not
been received, and
18
CA 02206870 1997-06-03
WO 96/18258 PCTIUS95/16391
the timeout period elapses, the user is prompted at step 1508 with a "Please
enter next digit"
message.
Once both digits for the memory location code are received and recognized, the
user
- receives a "Storing memory (two-digits), correct?" message, indicating the
two-digit memory
location number the system has collected. If the digits are incorrect, the
speaker-independent
recognizer is reengaged, the user is prompted at step 1512 with a "Location
please" message and
control returns to step 1504 to enable the user to reenter the memory location
digits. If the
received memory location number is correct, a test is performed at step 1516
to determine if the
storage memory is full. If not. the user receives a "Storing" message and the
number is stored in
FLASH memory at step 1518 at the indicated memory location number. System
control then
returns to step 304 of the originating call command handler routine of FIGURE
8.
If step 1516 determines the FLASH memory locator is full, .the user is
provided with a
"Memory full, erase?" message at step 1520. The speaker-independent recognizer
then monitors
at step 1522 for the voice command "ERASE." If no "ERASE" command is received,
the
procedure ends and control passes to step 304 of the originating call/command
handler routine of
FIGURE 8. If the "ERASE" command is detected, the user is prompted at step
1524 to confirm
the erase instruction, and the voice recognizer again monitors at step 1526
for the voice
command "ERASE." If the "ERASE" command is confirmed, the identified location
entry is
erased at step 1528.
Referring now to FIGURE 15, there is illustrated the directory speed dial
number storage
routine. This routine stores a number in a personal directory responsive to
speaker- dependent
directory names. The personal directory will be responsive only to commands
from a particular
user. Initially, the routine checks at step 1604 to determine if the selected
number is already
19
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
WU 96/1'/'J1
stored in FLASH memory. If the number is presently stored, the user is
prompted with a
"Re-enroll" message at step 1616. The speaker-independent speech recognizer
then monitors at
step 1618 to determine if the user wishes to re-enroll the directory name. If
the user wishes to
re-enroll the directory name, the user is prompted at step 1610 for a name to
be associated with
the number. If the directory name is not to be re-enrolled, an "Erase" message
prompts the user
and the voice recognizer monitor is engaged at step 1620. Inquiry step 1622
determines if the
user wishes to erase the directory number. If not, control passes to step 304
of FIGURE 8. If the
directory number is to be erased, the associated number is erased and two
beeps notify the user of
the erasure at step 1624. Command control then returns to step 304 of FIGURE
8.
If inquiry step 1604 determines the number is not presently enrolled in
memory, the user
is prompted with an "Enroll"-message and engages the speaker-independent
recognizer. A test at
step 1608 determines if the number should be enrolled. If the number is not to
be enrolled,
control passes to step 304 of the originating call/command handler routine of
FIGURE 8.
Otherwise, the user is prompted at step 1610 for a name to be associated with
the number and the
speaker-dependent recognizer is engaged. The user speaks the name to be
associated with the
number three times at step 1612 to program the speaker-dependent template. The
new directory
name and associated telephone number are stored at step 1614 in FLASH memory.
and the user
is notified of this by two beeps. Control then passes to step 304 of the
originating call/command
handler routine of FIGURE 8.
Referring now to FIGURE 16, there is illustrated the system access routine.
The system
access routine enables access to the platform either remotely beginning at
point 700 or locally
beginning at point 1300. The access pathway depends on whether a caller from
an outside line is
accessing the platform, or the, caller is using one of the local platform
telephones. When the
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
platform is accessed remotely, the outgoing message being played is
discontinued and the caller
is provided with a "Ready" message at step 702. The system then waits for a
command trigger
and sets the remote access flag. The user enters at step 704 a four-digit
speaker verification code
to access the platform. A test performed at step 706 determines if the
verification code is valid.
If not, the caller is notified of the invalidity of the code, the call is
disconnected and flags are
reset at step 1316.
When a valid code is received, a "Command please" message prompts the user at
step
708, and the platform waits for a command trigger. At inquiry step 710, the
platform monitors
for either the voice command "MESSAGES" or a DTMF "1" digit. If one of these
triggers are
detected, the system determines at inquiry step 1318 if any messages are
available. When no
messages are available,' control returns to step 708 to await additional
command triggers. If
messages exist, control passes to message playback routine of FIGURE 17.
Inquiry 712 tests for
either the voice command "FUNCTIONS" or a DTMF "2" digit. If either of these
triggers are
detected, control passes to the peripheral control routine of FIGURE 18.
Otherwise the system
generates two beeps to the remote caller at step 714 and control passes back
to step 708.
If a user accesses the system locally through point 1300, the user is
initially prompted
with a "Command please" message at step 1302 and awaits a command trigger.
Inquiry step
1304 determines if the voice command "MESSAGES" is spoken and, if so, the user
is prompted
at step 1310 to enter a verification code. The user enters a four digit
speaker verification code at
step 1312, and the validity of the code is tested at inquiry step 1314. If the
verification code is
valid. control passes to step 1318 to provide the user access to his messages
as previously
described. If the verification code is invalid, control passes to step 1316
and the user is notified
of the invalid code and disconnected from the platform. If no "MESSAGE''
command is
21
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
received, inquiry step 1306 monitors for a voice command "FUNCTIONS" and
passes control to
the peripheral control routine if the "FUNCTIONS" command is detected.
Otherwise, two beeps
are generated to the user at step 1308 and control passes back to step 1302 to
monitor for
command triggers.
Refernng now to FIGURE 17, there is illustrated the message playback routine.
This
routine enables a user to review recorded messages. Messages are retrieved at
step 1702 based
upon the speaker verification number provided by the user, and the messages
associated with the
identified speaker are played. Inquiry step 1704 monitors for receipt of DTMF
digits or any
spoken commands during playing of the messages. If neither is detected, step
1722 monitors for
the end of the selected messages and continues playback at step 1720 until the
end of the
messages is detected. Once the end of the messages is reached, the user is
prompted at step 1724
with a "Clear Messages" message. Inquiry step 1726 determines if the messages
are to be
cleared. If the messages are not to be cleared, control passes to step 1730
wherein the call is
disconnected if the remote access flag is set. If the messages are to be
cleared, they are erased at
step 1728. Control then returns to the executive routine of FIGURE 4.
If a DTMF digit or spoken command is detected at step 1704, inquiry step 1706
determines if the received command is either a DTMF "3" digit or a spoken
command "SKIP."
If so, the systems proceeds to the next recorded message at step 1704. If a
DTMF "4" digit or
"BACK" command is detected at inquiry step 1708, the current message is
repeated at step 1716.
If either a DTMF "5" digit or "DELETE'' command is detected at step 1710, the
current
message is erased at step 1718. If none of the above-described triggers are
detected, the input is
ignored at step 1712 and control passes back to step 1722.
22
CA 02206870 1997-06-03
WO 96/18258 PCT/US95/16391
Refernng now to FIGURE 18, there is illustrated the peripheral control
routine. The
peripheral control routine enables a user to access and control a variety of
systems within a
home or business through the telephone system using a series of voice
commands. When the
- peripheral control routine is accessed, the user is initially prompted at
step 1824 with a "Ready"
message. The system then monitors for the spoken command "ANSWER" at step
1802. If the
"ANSWER" command is received, the telephone answering device functions are
turned on or off
at step 1814. If no "ANSWER" spoken command is received, the system type is
checked at step
1804. Test step 1806 monitors for the command enhanced. If not detected,
control returns to the
prompting step 824. If the enhanced command is received, inquiry steps 1808,
1810 and 1812
monitor for the commands lights, thermostat or alarm respectively. If none of
these spoken
commands are detected, control returns to prompting step 1824.
If a "LIGHTS" command is received, the lights may be dimmed or turned on or
off at
step 1816 using spoken commands. If the "THERMOSTAT'' command is received, the
user may
check and change the thermostat settings using spoken commands at step 1818.
If the
"ALARM" command is received, the user may turn on or off the perimeter alarm
system using
spoken commands at step 1820. Once access of the peripheral devices is
complete, control
passes to step 1824 wherein the call is disconnected, remote access flag is
set and all other flags
cleared at step 1822 wherein all flags are reset and the call is disconnected.
The system then
returns to the executive routine described in FIGURE 4.
While the above description of FIGURES 4 through 18 have been given with
respect to
the use of vocal commands to enable access to the variety of telephone
services made available
to user by the intelligent call platform. the above described platform would
function equally well
23
CA 02206870 1997-06-03
WO 96118258 PCT/US95/16391
using pushbutton entry of DTMF digit commands as opposed to vocal commands.
Alternatively,
both voice and/or vocal commands could be utilized.
It should be appreciated by those skilled in the art that the specific
embodiments
disclosed above may be readily utilized as a basis for modifying or designing
other structures for
carrying out the same purposes of the present invention. It should also be
realized by those
skilled in the art that such equivalent constructions do not depart from the
spirit and scope of the
invention as set forth in the appended claims.
24
