 |
Claims  |
|
|
What is claimed is:
1. An automatic dial telephone by which a user can voice initiate a call to
a selected station and/or party, the automatic dial telephone comprising:
first means for storing a plurality of discrete voice reference patterns,
each discrete voice reference pattern representing a party to be called;
second means for storing a telephone number for each such party;
third means for storing a user voice reference pattern representing a
command for confirmation;
means for comparing a voice input pattern with discrete voice reference
patterns similar thereto to obtain a list of discrete voice reference
patterns similar to the voice input pattern;
means for outputting audible signals, representing the discrete voice
reference patterns from the list of discrete voice reference patterns, in
order of descending similarity to the voice input pattern, said outputting
means including pause means, to produce a pause period following each
audible signal output, for a user to input a user voice signal command for
confirmation of the identity of the party to be called;
said command for confirmation operating to stop said outputting means from
outputting audible signals representing any discrete voice reference
patterns remaining on the list of discrete voice reference patterns
following an output of an audible signal, representing a discrete voice
reference pattern from the list of discrete voice reference patterns,
which is followed by said command for confirmation; and
means for selecting and transmitting, as a signal, the telephone number of
the identified party to be called.
2. The telephone of claim 1, further comprising
control means for orienting the telephone to an off-hook condition for
reception thereby of a voice input which is denotive of a specific party
to be called;
said third means for storing having stored therein a user voice reference
pattern for a command for off-hook;
said means for comparing operating to recognize said user voice reference
pattern for said command for off-hook; and
means for operating the control means, responsive to said comparing means,
for causing said telephone to be placed off-hook.
3. The telephone of claim 1, further comprising means for repeating the
audible signal output.
4. The telephone of claim 1, further comprising means for generating a
signal upon failure to recognize a command for confirmation which
indicates such failure to the user.
5. The telephone of claim 1, as embodied in a motor vehicle, the motor
vehicle including vehicle operating systems,
the operating systems having normal and abnormal condition states,
the abnormal states being indicative of an accident potential operating
condition,
the telephone including means for detecting the abnormal states, and
the telephone including means for inhibiting the voice output during the
abnormal states.
6. The telephone of claim 1, in which said outputting means includes means
for producing a synthesized voice according to a recognition result, said
synthesized voice being effective to output ones of said plurality of
discrete voice reference patterns in descending order of similarity to the
voice input pattern.
7. The telephone of claim 1, further including means for storing voice
synthesis data for a plurality of parties to be called for voice synthesis
thereof.
8. The telephone of claim 1, further including means comprising a key input
portion for inputting a user command for confirmation of the identity of
the party to be called.
9. The telephone of claim 1 further comprising:
a child set for communicating between the user and a parent set,
the child set including a microphone for receiving voice inputs from the
user, a key for key signal inputs, and means for converting the voice and
key inputs into radio frequency signals for transmission to the parent
set,
the child set further including means for receiving radio frequency signals
from the parent set,
the parent set also including transmitting means and means for converting
the parent set radio frequency signals into voice signals to the user,
the parent set including means for receiving user inputs transmitted from
the child set, the parent set also including a keypad for receiving user
key signal inputs, and
the parent set further including means for controlling a telephone channel
and for communicating between the child set and parties to be called.
10. An automatic dial telephone by which a user can voice initiate a call
to a selected station and/or party, the automatic dial telephone
comprising:
first means for storing a plurality of discrete voice reference patterns,
each discrete voice reference pattern representing a party to be called;
second means for storing a telephone number for each such party;
third means for storing a user voice reference pattern representing a
command for confirmation;
means for comparing a voice input pattern, denotive of a specific party to
be called, with stored discrete voice reference patterns representing a
party to be called similar thereto;
means for outputting audible signals to the user indicative of parties
represented by stored discrete voice reference patterns in descending
order of similarity to the voice input pattern, the means for outputting
including pause producing means operable following each audible signal to
produce a pause period for a user to input a voice signal command for
confirmation of the identity of the party to be called;
means for comparing a voice input pattern, denotive of a specific user
voice signal command, with stored user voice reference patterns
representing user voice signal commands similar thereto;
means for determining the degree of similarity between said voice input
pattern, denotive of the specific party to be called, and of the specific
user voice signal command, respectively with each of the stored plurality
of discrete voice reference patterns representing parties to be called and
the stored user voice reference patterns representing user voice signal
commands, and including means for determining which of said voice
reference patterns has a degree of similarity to said voice input pattern
that is within a predetermined threshold;
means for selecting and transmitting as a signal the telephone number of
the identified party to be called when the user voice signal command for
confirmation is within the predetermined threshold; and
means for repeating comparison of the input voice pattern, denotive of the
specific party to be called, and of the specific user voice signal
command, respectively with the stored plurality of discrete voice
reference patterns representing parties to be called and the stored user
voice reference patterns representing user voice signal commands when the
degree of similarity is outside of the predetermined threshold.
11. An automatic dial telephone by which a user can voice initiate a call
to a selected station and/or party, the automatic dial telephone
comprising:
first means for storing a plurality of discrete voice reference patterns,
each discrete voice reference pattern representing a party to be called;
second means for storing a telephone number for each such party;
third means for storing a plurality of user voice reference patterns;
said plurality of user voice reference patterns representing a plurality of
commands;
said plurality of commands including commands for confirmation and for
off-hook;
means for orienting the telephone to an off-hook condition for reception
thereby of an input denotive of a specific party to be called, when a user
voice input pattern for off-hook is recognized;
means for placing the telephone to an on-hook condition, wherein the
telephone awaits an off-hook command, when a user voice signal is not
input within a first predetermined pause period following the off-hook
command;
means for comparing a voice input pattern, denotive of a specific party to
be called, with a stored plurality of discrete voice reference patterns
similar thereto when the telephone is in an off-hook state;
means for outputting an audible signal to the user indicative of a party
represented by a stored voice reference pattern which is most similar to
the voice input pattern, denotive of a specific party to be called, and
including pause producing means operable following the audible signal to
produce a second pause period for a user to input a voice signal command
for confirmation of the identity of the party to be called;
means for returning the telephone to an on-hook condition when the voice
signal command for confirmation is not input during the second pause
period; and
means for selecting and transmitting as a signal the telephone number of
the party to be called.
12. An automatic dial telephone by which a user can voice initiate a call
to a selected station and/or party, the automatic dial telephone
comprising:
first means for storing a plurality of discrete voice reference patterns,
each discrete voice reference pattern representing a party to be called;
second means for storing a telephone number for each such party;
third means for storing a plurality of user voice reference patterns;
said plurality of user voice reference patterns representing a plurality of
commands;
said plurality of commands including commands for confirmation and for
off-hook;
means for comparing a voice input pattern, denotive of a command, with a
stored plurality of user voice reference patterns;
said stored plurality of user voice reference patterns including a stored
voice reference pattern representing a command for off-hook;
means for enabling a control means when the means for comparing recognizes
a voice input pattern for off-hook as stored in the second means for
storing with the telephone in an on-hook condition, wherein the telephone
awaits an off-hook command;
means for orienting the telephone to an off-hook condition for reception
thereby of a voice input pattern denotive of a specific party to be
called;
first means for determining a degree of similarity between the stored user
voice reference pattern representing the command for off-hook and the
voice input pattern, including means for determining that the degree of
similarity is within a predetermined threshold;
the first means for determining including means for returning the telephone
to the on-hook condition when the degree of similarity is outside the
predetermined threshold;
means for comparing a voice input pattern, as an input denotive of a
specific party to be called, with a stored plurality of discrete voice
reference patterns similar thereto with the telephone in an off-hook
state;
means for outputting a plurality of audible signals to the user, indicative
of parties represented by stored discrete voice reference patterns, in
descending order of similarity to the voice input pattern;
said means for outputting including pause producing means effective,
following each one of said plurality of audible signals, to produce a
pause period for a user to input a voice signal command for confirmation
of the identity of the party to be called;
second means for determining the degree of similarity between the stored
user voice reference pattern, representing a command for confirmation, and
said voice signal command, including means for determining that the degree
of similarity is within a predetermined threshold;
means for selecting and transmitting as a signal the telephone number of
the identified party to be called; and
the second means for determining including means for returning the
telephone to the on-hook condition when the degree of similarity is
outside the predetermined threshold.
13. An automatic dial telephone comprising:
means for storing a first plurality of voice patterns each denotive of a
name of a party to be called and a related telephone number;
means for storing a second plurality of voice patterns each denotive of a
command;
said second plurality of voice patterns including at least an off-hook
command pattern and a confirmation command pattern;
means, effective when said telephone is on-hook, for comparing a voice
input pattern with said off-hook command pattern;
means responsive to a similarity of said voice input pattern and said
off-hook command pattern being within a predetermined threshold, for
placing said telephone off-hook;
means, effective when said telephone is off-hook, for comparing a voice
input pattern with said first plurality of voice patterns, and for
outputting a signal to a user indicative of a one of said first plurality
of voice patterns most similar to said voice input pattern;
means effective, following said outputting a signal, for comparing a
further voice input pattern with said confirmation command pattern and, if
said comparing indicates a similarity within a predetermined threshold,
for dialing said related number; and
means for placing said telephone on-hook if a similarity between said
further voice input pattern and said confirmation command pattern is
outside said predetermined threshold. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to automatic dial telephones, and
more particularly, to an automatic dial telephone for automatically
dialing numbers by speech recognition or a manual operation.
2. Description of the Prior Art
In recent years, with the improvement in recognition rate in a voice
recognition unit, a telephone capable of automatically dialing numbers in
respone to voice command has been put into practice (for example, Japanese
Patent Laid-Open Gazette No. 81152/1987).
For example, as a voice recognition processing method in such an automatic
dial telephone, a two-step recognition processing method has been
generally employed in which the name of a called party is recognized in
the first step and a command word such as "CANCEL" or "DIAL" is recognized
in the second step.
More specifically, the name of a called party spoken by a user is
recognized by speech recognition, and the results of this recognition are
displayed by visual or audible means. At this display, the user inputs a
command word for confirmation "DIAL" when the result of the recognition of
the name of the called party is correct. If this inputted voice is
recognized by the telephone as the command word for confirmation, a
telephone number of the called party is dialled.
On the other hand, when the result of the recognition of the name of the
called party is not correct, the user inputs a command word for
cancellation "CANCEL". If this inputted voice is recognized by the
telephone as the command word for cancellation, the automatic dial
telephone returns to a wait state in which the user must the name of the
called party.
The above described voice recognition processing method has the
disadvantage in that the user must input the command word for cancellation
and repeat the name of called party every time the name of the called
party is erroneously recognized, thereby imposing a large burden on the
user.
Such errors are liable to occur due to external noise because the name of
the called party, and the command words are inputted by voice. This
possibility is high, particularly if the automatic dial telephone is being
used in a vehicle such as a car.
SUMMARY OF THE INVENTION
An object of the present invention is to provide an automatic dial
telephone in which the burden on a user is reduced.
Another object of the present invention is to bypass relatively quickly any
error which may be due to external noise.
Still another object of the present invention is to provide an automatic
dial telephone that when used in a moving vehicle is capable of
automatically dialing numbers under safe driving conditions only.
Briefly stated, in an automatic dial telephone, that is also usable in a
motor vehicle, a user voice inputs a name of a party to be called. A voice
pattern of the name of the party to be called is compared with reference
patterns of parties to be called, stored in a reference pattern storage,
to determine the degree of similarity therebetween. The names of candidate
parties to be called are output to the user in order of decreasing
similarity. Each time the name of a called party is output, the automatic
dial telephone waits a predetermined time for the user to input a command
word for confirmation. When a voice confirmation command is input and
recognized during this waiting period, a telephone number corresponding to
the name of the called party is supplied to a telephone. Consequently, the
command word for confirmation may be input only if the name of the party
that is output is the one the user desires to call. Sensors continually
monitor the driving condition of the motor vehicle in which the telephone
is installed. When the operation of the steering wheel or brakes of the
motor vehicle exceeds a predetermined threshold or the speed of the motor
vehicle is excessive, the sensors generate safety signals that inhibit the
operation of the telephone.
According to an embodiment of the invention, there is provided an automatic
dial telephone by which a user can voice initiate a call to a selected
station and/or party, the automatic dial telephone comprises a memory for
storing a plurality of discrete voice reference patterns each discrete
voice reference pattern representing a party to be called. There is
another memory for storing a telephone number for each party to be called.
There is yet another memory for storing a user voice reference pattern
representing a command for confirmation. Also included is a device for
comparing a reference pattern of a name of a party to be called, which is
voice input, with the stored plurality of discrete voice reference
patterns similar thereto, to obtain a list of discrete voice reference
patterns similar to the voice input pattern. There is an output device for
outputting audible signals, representing the discrete voice reference
patterns from the list of discrete voice reference patterns, in order of
descending similarity to the voice input pattern. The output device
produces a pause following each audible signal that is output to allow a
user time to input a user voice signal command for confirmation of the
identity of the party to be called. Once the command for confirmation is
input, the output device stops outputting audible signals. Lastly, the
telephone of this embodiment has a device for selecting and transmitting,
as a signal, the telephone number of the identified party to be called.
According to a feature of the embodiment of the invention, there is also
provided a device for comparing a voice input pattern, denotive of a
specific user voice signal command, with the user voice reference pattern
similar thereto. Also included is a device for determining the degree of
similarity between the voice input pattern, denotive of the specific party
to be called, and of the specific user voice signal command, respectively
with each of the stored plurality of discrete voice reference patterns
representing parties to be called and the user voice reference pattern of
command words. The device for determining also determines which of the
voice reference patterns has a degree of similarity to the voice input
pattern that is within a predetermined threshold. When the degree of
similarity is within the predetermined threshold, a device selects and
transmits, as a signal, the telephone number of the identified party to be
called. When the degree of similarity is outside the predetermined
threshold, a device repeats the comparison of the input voice pattern and
the stored voice reference patterns.
According to a further feature of the embodiment of the invention, there is
also provided a plurality of commands including commands for confirmation
and for off-hook. Also included is a device for orienting the telephone to
an off-hook condition when a user voice input pattern for off-hook is
recognized. There is also a device for placing the telephone on-hook when
a user voice signal, denotive of a specific party to be called, is not
input within a first predetermined pause period following the off-hook
command. There is also a device for comparing a voice input pattern,
denotive of a specific party to be called, with the discrete voice
reference patterns similar thereto when the telephone is in an off-hook
state. Finally, there is a device for returning the telephone to an
on-hook condition when the user voice reference pattern command for
confirmation is not input during the second pause period, following the
output of the audible signal.
According to a still further feature of the embodiment of the invention,
there is also provided a device for comparing a voice input pattern,
denotive of a command, with the plurality of reference patterns. Also
included is a first determining device for determining a degree of
similarity between the stored voice reference pattern command for off-hook
and the voice input pattern, including a device for determining that the
degree of similarity is within a predetermined threshold. The first
determining device also includes a device for returning the telephone to
the on-hook condition when the degree of similarity is outside the
predetermined threshold. There is a second determining device for
determining the degree of similarity between the stored user voice
reference pattern, representing a command for confirmation, and the voice
signal command, including a device for determining that the degree of
similarity is within a predetermined threshold. The second determining
device also includes a device for returning the telephone to the on-hook
condition when the degree of similarity is outside the predetermined
threshold.
The foregoing and other objects, features, aspects and advantages of the
present invention will become more apparent from the following detailed
description of the present invention when taken in conjunction with the
accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram showing the electrical construction of a parent
set of a wireless automatic dial telephone;
FIG. 2 is a block diagram showing the electrical construction of a child
set of the wireless automatic dial telephone;
FIGS. 3A, 3B, 3C and 3D are flow charts showing the call by a control
portion of the parent set;
FIG. 4 is a flow chart showing the safety confirmation processing;
FIGS. 5A, 5B and 5C are flow charts showing the procedure of another call
processing;
FIGS. 6A and 6B are flow charts showing the procedure of still another call
processing; and
FIGS. 7A, 7B and 7C are flow charts showing the procedure of further call
processing.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIGS. 1 and 2 show the electrical construction of a wireless automatic dial
telephone, where FIG. 1 shows a parent set and FIG. 2 shows a child set.
Referring to FIG. 2, the child set comprises a microphone 1 for inputting a
voice, an amplification portion 2 for amplifying a voice signal from the
microphone 1, a transmission portion 3 comprising a modulator circuit for
modulating a carrier wave by the voice signal amplified and an amplifier
circuit for amplifying a modulated wave outputted from the modulator
circuit, an antenna coupler 4, an antenna 5 for transmission or reception,
a reception portion 6 including an amplifier circuit for amplifying the
modulated wave received and a demodulator circuit for demodulating the
modulated wave amplified, an amplification portion 7 for amplifying a
voice signal reproduced by the reception portion 6, a loudspeaker 8 for
outputting audibly the amplified voice signal, a key pad input portion 9
having a ten-key and various function keys, and a control portion 10 for
controlling the transmission portion 3, the reception portion 6 and the
key input portion 9.
Referring now to FIG. 1 parent set comprises a loop detection portion 27
for detecting the state of a telephone line loop, a voice recognition
portion 25 for automatically dialing numbers by voice input, a telephone
line control portion 26 for supplying to a telephone line a telephone
number corresponding to a voice recognized called party by the voice
recognition portion 25 and supplying to the telephone line a transmission
or reception signal, an antenna 21 for transmission or reception, an
antenna coupler 22, a reception portion 23 comprising an amplifier circuit
for amplifying a modulated wave received and a demodulator circuit for
demodulating the modulated wave amplified, a selecting switch 24 for
selecting either the telephone line control portion 26 the voice
recognition portion 25 to receive the voice signal reproduced by the
reception portion 23, a key input portion 34 having a ten-key pad and
various function keys, a voice synthesis portion 28 for generating various
synthetic voices, a transmission portion 29 including a modulator circuit
for modulating a carrier wave by an output signal of the voice synthesis
portion 28 or the reception signal from the telephone line control portion
26 and an amplifier circuit for amplifying a modulated wave from the
modulator circuit, and a control portion 30 for controlling these
equipments.
The speech recognition portion 25 comprises an analysis portion 41 for
analyzing the frequency spectrum of the voice signal inputted, a voice
pattern generation portion 42 for generating a group of time series of the
voice spectrum, that is, a voice pattern based on results of the analysis
by the analysis portion 41, an inputted voice pattern memory 45 for
storing the voice pattern generated by the voice pattern generation
portion 42, a first reference pattern memory 43 for storing reference
voice patterns of the names of a plurality of called parties, a second
reference pattern memory 44 for storing reference voice patterns of a
plurality of command words, and an identification portion 46 for comparing
the voice pattern outputted from the voice pattern generation portion 42
with the reference patterns to identify the degrees of similarity between
the patterns.
The control portion 30 controls a main memory 31 for storing a program and
required data, a recognition result memory 32 for storing the names of the
called parties in the order of decreasing degree of similarity in
accordance with speech recognition of the name of the called party, and a
counter 33 for measuring time.
Furthermore, a detection signal of a rotation angle detector 35 for
reporting the angle of rotation of a steering wheel of a car equipped with
this telephone, a detection signal of a brake detector 36 for reporting an
operating state of the brakes and a detection signal of a speed detector
37 for reporting the current speed of a car are inputted to the control
portion 30.
Before a user is able to automatically dial numbers by voice using this
wireless automatic dial telephone, the command words and the names of the
called parties must be trained into the telephone. The command words
comprise a command word for bringing the telephone into an off-hook state
"OFF-HOOK", a command word for configuration "OK" and the like.
The command words and the names of the called parties are trained into the
telephone in the following manner. First, a training key provided in the
key input portion 9 in the child set is depressed. A training operation
start signal is outputted from the control portion 10 in the child set.
The carrier wave is modulated by this training operation start signal in
the transmission portion 3, and the modulated wave is transmitted through
the coupler 4 and the antenna 5. This modulated wave is received by the
antenna 21 in the parent set, to be sent to the reception portion 23
through the coupler 22. In the reception portion 23, the modulated signal
received is demodulated and amplified. The training operation start signal
reproduced in the reception portion 23 is sent to the control portion 30,
initiating a training operation. As an alternate training key of the key
input portion 34 in the parent set may be to initiate training. In this
case, an output of the key input portion 34 is provided to the control
portion 30 in the parent set, initiating the training operation.
Then, the command words and the names of the called parties are inputted by
the microphone 1 of the child set. The voice signal outputted from the
microphone 1 is amplified in the amplification portion 2 and then, sent to
the transmission portion 3. In the transmission portion 3, the carrier
wave is modulated by the voice signal, and the modulated wave is
amplified. The modulated wave outputted from the transmission portion 3 is
transmitted through the coupler 4 and the antenna 5.
This modulated wave is received by the antenna 21 in the parent set, to be
sent to the reception portion 23 through the coupler 22. In the reception
portion 23, the modulated signal received is demodulated and amplified.
The voice signal reproduced in the reception portion 23 is sent to the
speech recognition portion 25 through the selecting switch 24.
In the speech recognition portion 25, the frequency spectrum of the voice
signal is first analyzed by the voice analysis portion 41. A voice pattern
is generated by the pattern generation portion 42, based on the results of
this analysis. This voice pattern is stored in the first reference pattern
memory 43 if the inputted voice is the name of a called party. The voice
pattern 75 stored in the second reference pattern memory 44 if the
inputted voice is a command word.
By such an operation, the voice patterns of the names of all the called
parties and the voice patterns of all the command words are stored in the
first or second reference pattern memories 43 or 44.
When training the names of called parties, a telephone number for each of
the called parties is inputted using the ten key pad of either key input 9
or 34 in the parent and child sets following the voice input of the name
of the called party, thus storing a telephone number with each of the
called parties trained in the main memory 31.
Referring now also to FIGS. 3A, 3B, 3C and 3D, to process a call using
control 30 the user turns the power supply of the child set on and then,
inputs a command word for off-hook "OFF-HOOK" using the microphone 1. When
the command word for off-hook is inputted to the microphone 1 (step 1),
the resultant voice signal outputted from the microphone 1 is amplified
and modulates a carrier wave. Then, the modulated wave is transmitted from
the antenna 5 in the child set.
This modulated wave is received by the antenna 21 in the parent set and
demodulated. The voice signal reproduced is sent to the voice recognition
portion 25 through the selecting switch 24 to perform voice recognition
processing (step 2).
In the voice recognition portion 25, the frequency spectrum of the voice
signal inputted is first analyzed by the voice analysis portion 41. A
voice pattern is generated by the pattern generation portion 42 based on
the results of this analysis. The voice pattern generated is stored in the
input pattern memory 45. In the identification portion 46, the voice
pattern generated is compared with reference patterns of all command words
stored in the second reference pattern memory 44, to identify the degrees
of similarity therebetween.
In the identification portion 46, when the inputted voice is recognized to
be "OFF-HOOK" (step 3), a direct-current loop is formed by the telephone
line control portion 26, resulting in an off-hook state (step 4).
Thereafter, safety confirmation processing is performed for examining
whether or not driving conditions are safe with respect to a telephone
operation performed by a user (step 5). The details of this processing
will be described later.
If it is confirmed that the driving conditions are safe during safety
confirmation processing, a synthetic voice signal representing a voice
guidance for urging the user to input the name of a called party, is
synthesized in the voice synthesis portion 28. The signal maybe, for
example, "input the name of the other party". This signal is sent to the
transmission portion 29, and the carrier wave is modulated by this signal.
The modulated wave is amplified and then, transmitted through the antenna
21.
This modulated wave is received by the antenna 5 in the child set, to be
demodulated. The synthetic voice signal reproduced is amplified and then,
sent to the loudspeaker 8. The voice for guidance is then outputted from
the loudspeaker 8 (step 6).
Thereafter, if the user inputs the name of the called party, for example,
"SANYO" at the microphone 1 (step 7), the voice signal is received in the
parent set by radio communication, and the degrees of similarity of this
voice signal to reference patterns of all called parties in the first
reference pattern memory 43 are calculated by the voice recognition
portion 25. The names of the called parties are stored in the recognition
result memory 32 in the order of decreasing degree of similarity (step 8).
Thereafter, a candidate order n is updated by one (step 9). The candidate
order n becomes one the first time because it is set to zero during
initialization.
Safety confirmation processing is then performed (step 10). If the safety
is confirmed a called party which is the n-th candidate of the called
parties stored in the recognition result memory 32 is read out. This
causes voice synthesis portion 28 to output a synthesized voice signal
representing the first candidate, which is the name of the stored called
party having the largest degree of similarity to the input called party.
This name is then outputted from the loudspeaker 8 in the child set (step
11).
Futhermore, a measuring operation for measuring predetermined time is
started (step 12). This predetermined time is set to a suitable time
period during which the user can input a command word for confirmation
"OK" with respect to voice output of the name of the called party. The
user inputs the command word for confirmation "OK" from that the
microphone 1 in the child set only if the name of the called party
outputted is the same as that inputted by the user.
If the user inputs the command word for confirmation "OK" from the
microphone 1 in the child set within the above described predetermined
time period (step 13), the voice signal is received in the parent set
through radio communication, and the inputted voice pattern is compared
with the reference patterns of all the command words in the second
reference pattern memory 44 by the voice recognition portion 25 (step 15).
If the inputted voice is recognized to be "OK" (step 16), a telephone
number corresponding to the name of the called party outputted in the
above described step 11 is read out from the main memory 31, to be sent to
the telephone through the telephone line control portion 26 (step 17).
Consequently, a phone call is made to the called party.
If the off-hook of the called party is detected by the loop detector 27
within a constant time period (step 18), the selecting switch 24 is
switched such that the voice signal inputted to the child set is supplied
to the telephone line control portion 26 (step 19). Consequently, a
telephone call to the called party is made. When the telephone call is
terminated (step 20), the contents of the main memory 31 are cleared and
the candidate order n is reset (n=0) (step 21), thereby terminating this
processing.
If a voice is not inputted within a predetermined time period after the
time measuring operation was started in the above described step 12 (step
14) and a voice is inputted but the inputted voice is not recognized to be
"OK" (step 16), it is determined whether or not the name of the other
party outputted in the above described step 11 is a final candidate (n=N)
(step 22).
If it is not the final candidate (n<N), the program is returned to the step
9. In the step 9, the candidate order n is updated by one. Thereafter,
safety confirmation processing is performed (step 23). If the safety is
confirmed, the name of a called party which is the n-th candidate, that
is, the next candidate is read out from the recognition result memory 32.
A synthetic voice signal corresponding to this called party is outputted
from the voice synthesis portion 28, and this synthetic voice is outputted
from the loudspeaker 8 in the child set on the basis of the synthetic
voice signal (step 11). In addition, the time measuring operation is
started (step 12).
The telephone is in a state in which it awaits the command word for
confirmation "OK" within a predetermined time period (steps 13 and 14). If
a voice is inputted during this wait state and the inputted voice is
recognized to be "OK" after the command for confirmation as a result of
the speech recognition (steps 13, 15 and 16), processing for automatic
dialing and a telephone call is performed (steps 17 to 21).
If a voice is not inputted in a state in which the command word for
confirmation "OK" is awaited (step 14) and a voice is inputted but the
inputted voice is not recognized to be the command word for confirmation
"OK" (step 16), it is determined whether or not the name of the other
party outputted this time is a final candidate (step 22). If the name of
the other party outputted this time is not the final candidate (n<N), the
program is returned to the step 9.
If the name of the called party outputted this time is the final candidate
(n=N), safety confirmation processing is performed (step 23) and then, a
voice for guidance asks the user whether or not to output the names of the
called parties again in the order of candidacy, for example, "Will you
output the name of the called party again?" is outputted (step 24). In
addition, predetermined time starts to be measured (step 25). This
predetermined time is set to a suitable time period during which the user
can input the command word for confirmation "OK" with respect to the
output of the voice for guidance. The user inputs the command word for
confirmation "OK" at the microphone 1 in the child set after the command
for confirmation only if it is desired to output the name of the called
party again.
If the user inputs the command word for confirmation "OK" within the above
described predetermined time period (step 26), its voice signal is
received in the parent set through radio communication, to be compared
with the reference patterns of the command words in the second reference
pattern memory 44 in the voice recognition portion 25 (step 28). If the
inputted voice is recognized to be "OK" (step 29), the candidat | | |
