U.S. patent application number 16/035895 was filed with the patent office on 2018-11-08 for system and method for processing speech files.
The applicant listed for this patent is NUANCE COMMUNICATIONS, INC.. Invention is credited to Julia HIRSCHBERG, Stephen WHITTAKER.
Application Number | 20180322186 16/035895 |
Document ID | / |
Family ID | 25373888 |
Filed Date | 2018-11-08 |
United States Patent
Application |
20180322186 |
Kind Code |
A1 |
HIRSCHBERG; Julia ; et
al. |
November 8, 2018 |
SYSTEM AND METHOD FOR PROCESSING SPEECH FILES
Abstract
A system and method for speech file processing which provides
users with differentially selectable speech file transcripts which
can be sent to one or more other users. The speech files may be
voicemail messages from which respective voicemail transcripts are
created. The voicemail transcripts are provided in a user
selectable format from which users may select non-contiguous
portions of the transcript.
Inventors: |
HIRSCHBERG; Julia;
(Cranford, NJ) ; WHITTAKER; Stephen; (Morristown,
NJ) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
NUANCE COMMUNICATIONS, INC. |
Burlington |
MA |
US |
|
|
Family ID: |
25373888 |
Appl. No.: |
16/035895 |
Filed: |
July 16, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15137781 |
Apr 25, 2016 |
10025848 |
|
|
16035895 |
|
|
|
|
14089839 |
Nov 26, 2013 |
9369581 |
|
|
15137781 |
|
|
|
|
13164287 |
Jun 20, 2011 |
8600745 |
|
|
14089839 |
|
|
|
|
12111533 |
Apr 29, 2008 |
7966181 |
|
|
13164287 |
|
|
|
|
11697964 |
Apr 9, 2007 |
7366665 |
|
|
12111533 |
|
|
|
|
09879313 |
Jun 12, 2001 |
7225126 |
|
|
11697964 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04M 2201/40 20130101;
H04L 51/36 20130101; G10L 15/26 20130101; H04M 3/533 20130101; H04M
2203/301 20130101; H04L 51/08 20130101; G06F 16/316 20190101; G06F
16/345 20190101 |
International
Class: |
G06F 17/30 20060101
G06F017/30; H04L 12/58 20060101 H04L012/58; G10L 15/26 20060101
G10L015/26; H04M 3/533 20060101 H04M003/533 |
Claims
1. A method comprising: displaying, at a graphical user interface,
a transcript of an audio file; receiving, via the graphical user
interface, a selection of a portion of the transcript of the audio
file; identifying the portion of the transcript selected based on a
transcript index which indexes text in the transcript relative to
an occurrence of the text in the audio file; and transmitting the
portion of the transcript selected to a particular recipient.
2. The method of claim 1, further comprising: receiving a second
selection of a second portion of the transcript associated with the
audio file, wherein the portion and the second portion are not
contiguous; and transmitting the second selection of the second
portion of the transcript the particular recipient.
3. The method of claim 1, wherein information associated with the
audio file is used to identify the particular recipient.
4. The method of claim 3, wherein the information comprises a name
associated with the audio file, a message summary, a telephone
number associated with the audio file, a date associated with the
audio file, a size of the audio file, or any combination
thereof.
5. The method of claim 1, further comprising: receiving an
indication for including the audio file with the portion of the
transcript with the portion of the transcript transmitted to the
particular recipient; and transmitting the indication for including
the audio file to a server for delivery of the audio file to the
particular recipient with the portion of the transcript.
6. The method of claim 5, wherein the audio file comprises only a
second portion of the audio file that corresponds to the portion of
the transcript.
7. The method of claim 1, wherein the transmitting uses an
electronic mail message, and wherein the electronic mail message
comprises an attachment of the audio file corresponding to the
portion of the transcript.
8. A communication device, comprising: a processor; and a memory
that stores executable instructions that, when executed by the
processor, facilitate performance of operations, comprising:
displaying, at a graphical user interface, a transcript of an audio
file; receiving, via the graphical user interface, a selection of a
portion of the transcript of the audio file; identifying the
portion of the transcript selected based on a transcript index
which indexes text in the transcript relative to an occurrence of
the text in the audio file; and transmitting the portion of the
transcript selected to a particular recipient.
9. The communication device of claim 8, wherein the memory stores
additional executable instructions that, when executed by the
processor, facilitate performance of operations further comprising:
receiving a second selection of a second portion of the transcript
associated with the audio file, wherein the portion and the second
portion are not contiguous; and transmitting the second selection
of the second portion of the transcript the particular
recipient.
10. The communication device of claim 8, wherein information
associated with the audio file is used to identify the particular
recipient.
11. The communication device of claim 10, wherein the information
comprises a name associated with the audio file, a message summary,
a telephone number associated with the audio file, a date
associated with the audio file, a size of the audio file, or any
combination thereof.
12. The communication device of claim 8, wherein the memory stores
additional executable instructions that, when executed by the
processor, facilitate performance of operations further comprising:
receiving an indication for including the audio file with the
portion of the transcript with the portion of the transcript
transmitted to the particular recipient; and transmitting the
indication for including the audio file to a server for delivery of
the audio file to the particular recipient with the portion of the
transcript.
13. The communication device of claim 12, wherein the audio file
comprises only a second portion of the audio file that corresponds
to the portion of the transcript.
14. The communication device of claim 8, wherein the transmitting
uses an electronic mail message, and wherein the electronic mail
message comprises an attachment of the audio file corresponding to
the portion of the transcript.
15. A non-transitory machine-readable storage medium comprising
executable instructions that, when executed by a processor,
facilitate performance of operations comprising: displaying, at a
graphical user interface, a transcript of an audio file; receiving,
via the graphical user interface, a selection of a portion of the
transcript of the audio file; identifying the portion of the
transcript selected based on a transcript index which indexes text
in the transcript relative to an occurrence of the text in the
audio file; and transmitting the portion of the transcript selected
to a particular recipient.
16. The non-transitory machine-readable storage medium of claim 15,
further comprising: receiving a second selection of a second
portion of the transcript associated with the audio file, wherein
the portion and the second portion are not contiguous; and
transmitting the second selection of the second portion of the
transcript the particular recipient.
17. The non-transitory machine-readable storage medium of claim 15,
wherein information associated with the audio file is used to
identify the particular recipient.
18. The non-transitory machine-readable storage medium of claim 17,
wherein the information comprises a name associated with the audio
file, a message summary, a telephone number associated with the
audio file, a date associated with the audio file, a size of the
audio file, or any combination thereof.
19. The non-transitory machine-readable storage medium of claim 15,
wherein the non-transitory machine-readable storage medium stores
additional executable instructions that, when executed by the
processor, facilitate performance of operations further comprising:
receiving an indication for including the audio file with the
portion of the transcript with the portion of the transcript
transmitted to the particular recipient; and transmitting the
indication for including the audio file to a server for delivery of
the audio file to the particular recipient with the portion of the
transcript.
20. The non-transitory machine-readable storage medium of claim 19,
wherein the audio file comprises only a second portion of the audio
file that corresponds to the portion of the transcript.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 15/137,781, filed Apr. 25, 2016, which is a
continuation of and claims priority to U.S. patent application Ser.
No. 14/089,839, filed Nov. 26, 2013, now U.S. Pat. No. 9,369,581,
issued Jun. 14, 2016 by Hirschberg et al., entitled, "System and
Method for Processing Speech Files", which is a continuation of and
claims priority to U.S. patent application Ser. No. 13/164,287,
filed Jun. 20, 2011, now U.S. Pat. No. 8,600,745, issued on Dec. 3,
2013, which is a continuation of and claims priority to U.S. patent
application Ser. No. 12/111,533, filed Apr. 29, 2008, now U.S. Pat.
No. 7,966,181, issued Jun. 21, 2011, which is a continuation of and
claims priority to U.S. patent application Ser. No. 11/697,964,
filed Apr. 9, 2007, now U.S. Pat. No. 7,366,665, issued on Apr. 29,
2008, which is a continuation of and claims priority to U.S. patent
application Ser. No. 09/879,313, filed Jun. 12, 2001, now U.S. Pat.
No. 7,225,126, issued on May 29, 2007. The contents of each of the
foregoing are hereby incorporated by reference into this
application as if set forth herein in full.
BACKGROUND OF THE DISCLOSURE
[0002] The present invention relates to the field of communications
and more particularly to a system that allows users to select one
or more portions of a speech file transcript and then provide only
the selected portions to one or more entities one or more
electronic formats.
[0003] Electronic mail and voicemail systems form the foundation of
corporate and personal communications. Electronic mail has proven
to be even more popular in recent years as the electronic mail
systems have become more robust with a variety of useful features
like electronic mail return receipts and the ability to attach and
transfer files along with the electronic mail messages. Some
current hybrid systems also have partially merged the two systems,
for example, by allowing a user to check their voicemail through
their electronic mail account. This is typically performed by
creating an electronic mail version of a voicemail message, such as
by having an electronic mail message with a digitized version of
the voicemail message attached to the electronic mail message. In
this case, the voicemail message may be stored and organized as
with other conventional electronic mail messages.
[0004] However, none of these prior art systems, electronic mail,
voicemail or any hybrid system allows users to selectively capture
portions of a voicemail message and forward or send only the
selected portions to one or more other users.
[0005] Accordingly, it would be desirable to have a system which
allows a user to select certain portions of a speech file, such as
a voicemail message, in an intuitive manner and then share such
selected portions with certain designated parties the user
specifies. It would be further desirable to have such an
aforementioned system which allows the user to select
non-contiguous portions of the speech file and then have only the
non-contiguous portions provided to parties the user has
specified.
SUMMARY OF THE INVENTION
[0006] The present invention is a system and method for processing
speech files to allow for selection of one or more portions of the
speech file for provision of only the selected portions to one or
more parties. In one embodiment of the present invention, the
method for processing voicemail messages includes the steps of
transcribing a plurality of voicemail messages to produce a
plurality of voicemail message transcripts, indexing the plurality
of voicemail messages transcripts, providing the voicemail message
transcripts to one or more users, receiving at least one selection
action from one or more of the users, the at least one selection
action identifying at least a portion of one or more of the
voicemail message transcripts for delivery to one or more parties
identified by the one or more users and providing the selected
portion of the one or more voicemail message transcripts to the one
or more parties specified by the one or more users.
[0007] The present invention includes a graphical user interface
for use in browsing, searching and selecting certain portions of
the speech files. The graphical user interface facilitates the
user's navigation of the user's messages to enable the same person
to have access to and the ability to search for information
contained in their voicemails and/or electronic mail messages. The
user interface may include a window or screen where the transcribed
text of the voicemail messages are displayed. Certain message
information such as the name of the caller, date of the call and
time of the call can be displayed in a separate window or
screen.
BRIEF DESCRIPTION OF THE DRAWINGS
[0008] FIG. 1 illustrates a messaging system in accordance with the
teachings of the present invention.
[0009] FIG. 2 illustrates an exemplary voicemail message server in
accordance with the teachings of the present invention.
[0010] FIG. 3 illustrates an exemplary transcript index in
accordance with the teachings of the present invention.
[0011] FIG. 4 illustrates an exemplary screen display in accordance
with the teachings of the present invention.
[0012] FIG. 5 illustrates an exemplary method in accordance with
the teachings of the present invention.
DETAILED DESCRIPTION OF THE DRAWINGS
[0013] Referring to FIG. 1, a messaging system 10 is illustrated in
FIG. 1 which enables users to view and select selective portions or
precis of speech based information and then provide such selected
information to certain parties identified by the users. In one
embodiment, messaging system 10 has at least one voice mail server
20. While the preferred embodiment of the present invention is
described and illustrated below as a messaging system having one
voice mail server, the present invention may easily be implemented
with two or more voice mail servers which may be in communication
with one another. In this manner, voice mail server 20 may be
connected via an inter-mailbox data network to other respective
voice mail servers, not shown, in messaging system 10. Thus, each
voice mail server would be able to communicate (e.g., transmit and
receive information) with the other voice mail servers in the voice
mailbox network system.
[0014] Referring again to FIG. 1, messaging system 10 is
illustrated as having voice mail server 20 connected as part of a
primary communications network 30, such as an intra company voice
mail system. It is understood that primary communications network
30 could be a private branch exchange (PBX), Centrex, or similar
communication or telecommunication system that controls access to
the voice mail server 20. The primary communications network 30
connects subscribers, such as subscribers 50 and 60, in the network
to the voice mail server 20.
[0015] In this embodiment, voice mail server 20 includes at least
one database 40, for storing, for example, voice mail message files
and voice mail message transcripts as discussed in more detail
later herein, as well as the operating programs for the particular
voice mail server served by database 40. Database 40 may be any
type or combination of types of storage media such as magnetic,
optical, optical-magnetic, etc. so long as the storage facility has
sufficient capacity to store a plurality of voice mail messages
from a plurality of subscribers.
[0016] In one embodiment, voice mail server 20 is preferably a
computer system that essentially functions as a central answering
machine for subscribers to the voice mail system. It is understood
that the present invention can be utilized in or adapted to a
variety of voice mail servers or similar equipment.
[0017] Voice mail server 20 is also connected via respective trunk
lines, not shown, to a communications network 70, which is
illustrated in FIG. 1 as preferably being the public switched
telephone network. In this manner, a caller may access the voice
mail server 20 via communications network 70 through use of a
portable telephone 80 and/or personal computer 90 or other similar
device. It is also understood that access to the voice mail server
is not intended to be limited to telephones and/or personal
computers, but could be, for instance, wireless devices,
conventional facsimile machines, palmtops, or any other device that
is capable of transmitting and receiving data over a telephone
line.
[0018] In the present embodiment, voice mail server 20 is in
communication with a message server 90, such as an electronic mail
message server, for user in delivering messages, such as certain
selections of voicemail transcripts and corresponding audio to one
or more entities. As discussed in more detail later herein, voice
mail server 20 processes speech files, in this case, the speech
files are voicemail messages, to produce one or more voicemail
transcripts. Users are then provided the opportunity to select one
or more portions of a voicemail transcript. The one or more
selected portions are provided to one or more identified recipients
via message server 90.
[0019] Referring to FIG. 2, a more detailed view of voicemail
server 20 is shown. In this embodiment, voicemail server 20
includes a message database 100, an automatic speech recognition
component 102, a message indexing component 106, a user selection
processing component 110 and a selection delivery component 114.
Typically, message database 100 receives and stores speech files,
such as voicemail messages. Automatic speech recognition is
performed upon these speech files by automatic speech recognition
component 102 to produce transcripts of the speech files. The
transcripts are then indexed by message indexing component 106 to
produce a transcript index, such as shown in FIG. 3, wherein each
word in the transcript is indexed relative to the occurrence of the
word in the speech file. In this manner, as discussed in more
detail later herein, selection of one or more words or portions of
the transcripts is easily identifiable based on the indexing. User
selection component 110 provides users the ability to select one or
more portions of the speech transcripts. The portions selected may
be non-contiguous, such as one word from one sentence, a few words
from another sentence, a phone number from another section of the
transcript, and other variations thereof. Once a portion or
portions of a speech transcript are selected, certain desired
recipients may be provided the selected portion or portions via
selection delivery component 114. In one embodiment, selection
delivery component 114 is an interface with a message server, such
as an electronic mail message server shown in FIG. 1, which
provides the selected portion or portions of a speech transcript to
one or more entities as electronic mail message(s).
[0020] Referring to FIG. 4, an exemplary user interface screen 200
as may be provided via selection delivery component 114 (FIG. 2) to
one or more users is shown. Screen 200 includes a message summary
section 210 and a message transcript section 220. In one
embodiment, message summary section 210 provides information such
as the name of the caller/sender, the size of the message, the
subject/telephone number, the date and other related message
information. Messages may be selected within message summary
section 210, such as highlighted message 230, which provides the
corresponding message transcript within message transcript section
220. From within message transcript section 220, text of the
message selected within message summary section 210 is provided. A
portion or portions of the transcript text may be selected within
message transcript section 220, such as shown by selected
non-contiguous portions 240, 244 and 248.
[0021] Referring to FIG. 5, an exemplary embodiment of a method for
processing speech files in accordance with the present invention is
shown. In this embodiment, one or more speech file(s) are received,
such as via a voicemail server discussed earlier herein, step 310.
Automatic speech recognition is performed on such speech file(s),
such a via an automatic speech recognition component discussed
earlier herein, step 320. The speech file(s) are indexed, such as
shown in FIG. 3, step 330. A transcript of the indexed speech
file(s) is provided to a user, such as shown in FIG. 4, step 340.
The user's selection of one or more portion(s) of the speech
file(s) transcript is received, such as also shown previously in
FIG. 4, step 350. The selected portion(s) of speech file transcript
is provided to one or more entities or parties specified by user,
step 360, such as via selection delivery component, discussed
earlier herein. In one embodiment, the entities or parties may
simply be electronic mail addresses or user names specified by the
user to which the selected portion or portions of the transcript
will be provided to. The specified recipients of the transcript
portion or portions may receive the portions in both a textual and
an audible format. For example, the portion or portions selected
may be provided as text within an electronic mail message with an
attachment of an audio file which corresponds to the selected
portion or portions.
[0022] In an exemplary embodiment, automatic speech recognition or
simply, speech to text techniques are used to derive text from
speech, i.e. to identify the letters or words spoken by a human
subject in one or more speech files, such as voicemail messages. In
the present invention, automatic speech recognition is used to
analyze the speech signals contained in a speech file, such as a
voicemail message to produce a textual transcript of the speech
signals in the voicemail message. In an exemplary embodiment, such
speech recognition techniques may use a combination of pattern
recognition and sophisticated guessing based on some linguistic and
contextual knowledge to transcribe the speech files. It is
contemplated that other methodologies and techniques may be used so
long as the speech is properly transcribed into a textual format to
produce a workable transcript from which a user may select one or
more portions from to send or forward on to one or more other
parties or entities.
[0023] In the present invention, transcribing of the voicemails by
automatic speech recognition is preferably performed automatically,
for example, as soon as a voicemail message is left for a user or
alternatively, transcribing may be performed periodically as
determined by the user or by system defaults. In one embodiment of
the present invention, automatic speech recognition is performed in
conjunction with or immediately subsequent to the recording of the
voice or speech signals as voicemail messages. For example,
transcribing may be performed as someone is leaving a voicemail
message by transmitting the voice signals to the respective
voicemail server for processing. Alternatively, transcribing may
performed immediately after the voicemail is saved on the voicemail
server by having the voicemail server first transmit the stored
voicemail message to the speech recognition component of the
voicemail server and then using automatic speech recognition to
transcribe the voicemail.
[0024] Alternatively, the system may wait until a certain
predetermined number of voicemails are stored for a certain user on
the voicemail server before processing the voicemails. Once the
certain predetermined number of voicemails is attained, processing
of the voicemail messages may be performed on the group of
voicemails by the speech recognition component. For example, the
system may be configured to transcribe voicemail messages after at
least two or more messages are left in a user's mailbox. As a
further alternative, transcribing of the voicemails may be
performed only after a user has actively selected for transcribing
to be performed on the voicemails. For example, the user may be
provided in the system with a menu selection or selection key which
when pressed or selected, would initiate transcribing of their
voicemails. The user may also be provided with the choice of having
specific voicemails of their choosing processed by the system. In
this instance, some users may prefer to listen to some of their
voicemails in the conventional manner while having other
voicemails, such as relatively longer voicemails, transcribed and
indexed by the system. It is contemplated that the system may
provide the user with the choice of having his/her voicemails
processed by the system. In one embodiment, the user may be charged
a certain fee for voicemail processing or alternatively, the
voicemail processing may be offered as a free value added
service.
[0025] Once the voicemails have been transcribed, the text of the
voicemail message(s) may be indexed using full text
indexing/retrieval techniques as known in the art. Once a user
selects a portion or portions of a speech file transcript as
described earlier herein, those selected portions are used in
conjunction with the transcript index, such as the one shown in
FIG. 3, to create a corresponding audio file containing only those
selected portion to provide to the one or more parties the user has
specified. In other words, the selected portions of the transcripts
which the user has selected are extracted from the original speech
file to produce a new speech file containing only the selected
portions. It is contemplated that any number of indexing/retrieval
techniques may be employed within the present invention to provide
for more efficient and faster information retrieval of selected
portions of the speech file transcripts.
[0026] In another embodiment of the present invention, a sound or
audio file of the voicemail message is also provided to the one or
more users. In one embodiment, the sound or audio file may be
provided as an attachment to the electronic mail message. The sound
or audio file may be provided as an MPEG-x Audio Layer-x (mpx) file
such as an mp3 file, a .WAV file, a streaming audio file or other
similar file format.
[0027] It will be apparent to those skilled in the art that many
changes and substitutions can be made to the system and method
described herein without departing from the spirit and scope of the
invention as defined by the appended claims.
* * * * *