CN100353417C - Method and apparatus for providing text message - Google Patents

Abstract

A method and apparatus for providing a text message includes an input of an electronic device (100) receiving an utterance (step 210). Speech recognition is then performed on the utterance (step 230) guided by a message template set by the user, where the message template is stored in a memory (155) associated with the electronic device (100). Speech recognition is defined by matching one of the utterances to a template to create a matching template. A text message is then provided through the matching template (step 235).

Description

Method and apparatus for providing text messages

technical field

The present invention generally relates to a method and apparatus for providing text messages using voice. The present invention is particularly useful, but not exclusively, for providing text messages using voice input that is processed on portable electronic devices with limited memory and computing power.

Background technique

Short text messages, usually in Short Message Service (SMS) format, are widely used in the field of wireless communications. Billions of short text messages are sent every month, often from one mobile phone to another. There are many reasons for the popularity of this text message. Messages typically cost a fraction of the cost of a minute mobile phone call and require no busy tone to be sent or received. Therefore, the sending user can create and send the message when it is convenient, and the receiving user can receive and read the message when it is convenient.

Text messages are typically created by typing characters into the keypad of the mobile phone. However, composing a message using such a small, non-standard keyboard can be inconvenient and often requires more time than with a full-sized standard keyboard. However, it is of course impractical to attach a full-sized standard keyboard to a mobile phone. Therefore, there is a need for a more efficient method of composing short text messages.

Furthermore, although there are various known speech recognition systems, most are not suitable for portable electronic devices such as mobile phones. This is because prior art speech recognition systems typically require more processing power and memory than can be provided in portable electronic devices.

Closed-vocabulary speech recognition systems and methods in the prior art use a pre-set, fixed vocabulary. In use, a fixed vocabulary may be large but not complete, so that, for example, it may not include a person's last name or many place names. In contrast, open vocabulary speech recognition systems and methods use a variable vocabulary to which a user can add new words and phrases or delete words and phrases. However, existing open-vocabulary speech recognition systems and methods require relatively high computational overhead, which may be prohibitive for portable electronic devices such as digital personal assistants (PDAs), wireless phones, and other portable devices.

In this specification and claims, the term "comprises" or similar terms is intended to indicate a non-exclusive inclusion, so that a method or apparatus comprising a group of components does not only include these components, but may well include other components. Components not listed.

Contents of the invention

According to one aspect of the present invention, a method for providing a text message is provided. The method includes the step of receiving an utterance at an input of an electronic device. performing speech recognition on the utterance guided by a message template set by the user, the message template being stored in a memory associated with the electronic device, wherein the speech recognition is defined by matching the utterance to one of the templates, to create a matching template. Then, provide a text message through this matching template.

At least one message template may contain fixed language components.

At least one message template may contain variable language components.

At least one message template can contain both fixed and variable language components.

Text messages may be SMS messages.

The above method may also include the step of editing the message template set by the user by receiving characters typed by the keyboard of the electronic device.

A portion of a text message may be a transcription of an utterance.

Text messages may all be transcriptions of utterances.

According to another aspect of the present invention, an electronic device for providing text messages is provided. The electronic device includes a microphone for receiving an utterance; a non-volatile memory for storing a message template; a processor for performing speech recognition of an utterance guided by the message template, wherein the processor interprets one of the utterance and the template Matches are made to build a match template through which the text message is served.

For the electronic device, the message template may also include fixed and/or variable language components.

For the electronic device, the text message may be an SMS message.

The electronic device may include a keypad for editing message templates.

The electronic device can match the utterance to the plurality of templates to calculate a likelihood for each template.

Description of drawings

In order that the present invention may be readily understood and carried into practice, reference will now be made to the preferred embodiments shown, with reference to the accompanying drawings. In the attached picture:

Figure 1 is a schematic block diagram of a radiotelephone according to the present invention.

FIG. 2 is a flowchart illustrating a method for providing, editing and sending a text message according to the present invention.

FIG. 3 is a flowchart illustrating a method for providing a list of alternative message templates according to the present invention.

4 is a flowchart illustrating a method of enabling a user to edit existing message templates and save new templates in static programmable memory in accordance with the present invention.

Detailed ways

Referring to FIG. 1, a radiotelephone 100 is shown which includes a radio frequency communication unit 105 connected to and in communication with a processor 110. An input/output (I/O) interface in the form of a display screen 115, a keyboard 120, a microphone 190, and a speaker 195 are also connected to and in communication with the processor 110.

Processor 110 includes encoder/decoder 125 and associated read-only memory (ROM) 130 storing data for encoding and decoding voice or other signals that may be transmitted or received by radiotelephone 100 . Processor 110 also includes microprocessor 135, random access memory (RAM) 150, static programmable memory 155, and removable SIM module 160, wherein microprocessor 135 is connected to encoder/decoder via common data and address bus 140 125 and its associated character read only memory (ROM) 145. Both static programmable memory 155 and SIM module 160 can store (among other messages) selected templates for incoming text messages, phonebook databases, and outgoing text messages (described in more detail below). Microprocessor 135 has ports for connection to keyboard 120, display screen 115, and prompt module 165, wherein prompt module 165 typically contains a speaker, vibration motor and corresponding drivers. Character ROM 145 stores codes for decoding or encoding text messages that may be received by communication unit 105 and entered on keyboard 120 .

The RF communication unit 105 is a combined receiver and transmitter with a common antenna 170 . The communication unit 105 has a transceiver 175 connected to an antenna 170 via a radio frequency amplifier 180 . The transceiver 175 is also connected to a combined modulator/demodulator 185, which connects the communication unit 2 to the processor 110.

Referring to FIG. 2 , a flowchart illustrates an embodiment of the present invention including a method 200 for providing, editing and sending text messages using the wireless telephone 100 . Method 200 is called in start step 205 . In step 210 utterance is received on an input device of phone 100 such as microphone 190 . Processor 110 then samples and digitizes the utterance waveform in step 215, then segments it in step 220, and then processes it in step 225 to provide feature vectors representing the waveform. It should be noted that step 215, step 220, and step 225 are well known in the art, and thus a detailed description thereof is not required.

Next, in step 230 , speech recognition is performed on the feature vector generated in step 225 . The speech recognition is guided by user-set message templates stored in the static programmable memory 155 of the device 100 . Message templates are described in detail below. The method 200 then provides a text message to the user in step 235 . The message may be provided to the user by using one of the I/O interfaces, such as display 115 or speaker 195 of device 100 .

After providing the message to the user, the user can then decide in step 240 whether to edit the message. If the user decides not to edit the message, the message is sent in step 245 in a message format such as SMS. However, if the user decides to edit the message at step 240, the message is edited at step 250 before it is sent at step 245. In various embodiments of the invention, the user may edit the message in several different ways, including voice-based editing through the speaker 195 or typing-in editing through the keyboard 120 . Method 200 then ends at step 255 .

In another embodiment of the present invention, after the speech recognition step 230 described above, the step 235 of providing a text message may include providing the user of the phone 100 with a list of alternative message templates from which the user can select the most suitable template for the desired message. template for your text message.

FIG. 3 is a flowchart illustrating a method 300 for providing such a list of alternative templates to a user. Method 300 is invoked in start step 305 when a user enters a command into keyboard 120 or microphone 190 . First, the method 300 includes the processor 110 selecting a message template from a list of available message templates in step 310 . Then, in step 315 , this selected template is compared to the feature vector provided in step 225 of method 200 . Next, the processor 110 calculates a likelihood value in step 320 , estimating the matching degree between various aspects of the selected template and the feature vector of the input utterance.

Then, the processor 110 determines in step 325 whether the likelihood value is higher than a preset threshold. This threshold can be calculated automatically by the processor 110 or preset by the user of the phone 100 . If the likelihood value of the selected template is lower than the preset threshold, the template is rejected in step 330 . However, if the likelihood value of the selected template is higher than the preset threshold, then in step 335, it is considered that the template has a reasonable matching degree with the input utterance, and it is added to the list of candidate templates. Regardless of whether the selected template is rejected or added to the candidate template list, method 300 proceeds to step 340 where processor 110 determines whether all available templates have been evaluated. If not all available templates have been evaluated, method 300 picks the next message template at step 345 and returns to step 315 where the next template is compared to the feature vector of the input utterance. If all templates have been evaluated in step 340, method 300 proceeds to step 350 and provides the user with a list of all candidate templates. Alternate templates may be presented to the user by using one of the I/O interfaces such as display 115 or speaker 195 of device 100 . Method 300 then ends at step 355 .

According to one embodiment of the invention, the user of the phone 100 is not limited to using only templates provided by the manufacturer of the device 100 . Instead, a user of device 100 may edit existing templates stored in static programmable memory 155 to create his own personalized message templates.

Referring to FIG. 4 , a method 400 of enabling a user to edit existing templates and save new templates in static programmable memory 155 is shown. Method 400 is invoked in start step 405 when a user enters a command into keyboard 120 or microphone 190 . In step 410 , a user of device 100 is provided with a list of existing templates through an I/O interface such as display screen 115 or speaker 195 . Next, the user selects a desired message template by using an I/O interface such as microphone 190 or keyboard 120 in step 415 . Next, in step 420 , the user edits the template again by using an I/O interface such as the microphone 190 or the keyboard 120 . Finally, in step 425 , the user saves the edited template in the static programmable memory 155 . Then, the method 400 ends at step 430 .

Other methods of editing message templates are also within the scope of the present invention, including by connecting the phone 100 to a host using a communication channel such as a USB cable, and subsequently downloading or flashing the edited template to static programmable memory 155.

The method of the present invention may further comprise message templates comprising fixed and variable language components. Fixed language components do not change when a user selects a template and sends a message. Variable language components can be modified by the user from message to message. The use of fixed and variable speech components greatly improves the leverage of the phone 100's limited processing power and memory.

For example, a specific template for a short text message related to a meet and greet request might look like this: " Meet me at $PLACE at $TIME.". In this sentence, fixed language components are underlined, while variable language components are capitalized and start with "$". Different users of the template can then edit variables such as $PLACE to suit their particular environment. For example, a college student might set the variable $PLACE to: $PLACE = sp|library(library)|dormitory(dormitory)|cafeteria(cafeteria), etc. and a lawyer might set the variable $PLACE to: $PLACE=sp|office (office)|courthouse (court)|home (home) etc.

In the above sentence, "sp" refers to pause or silence, and "|" refers to the logical operator "OR".

Another example of a message template that may be used in the present invention is: " Happy $FESTIVAL.". The variable linguistic component $FESTIVAL in this sentence might be edited by the user to:

$FESTIVAL＝sp|birthday(birthday)|new year(New Year)|thanksgiving(Thanksgiving) etc.

Using open-lexical speech recognition, the phone 100 is able to recognize edited variable language components entered by the user. Since the variable linguistic components consist of discrete sets of variables, the overhead and memory requirements of speech recognition processing are minimized. The method described above is therefore particularly suitable for devices with only limited processing and memory resources, such as mobile phones.

The use of templates comprising fixed and variable linguistic components increases the efficiency of speech recognition systems in several ways. First, since the number of templates stored in static programmable memory 155 is modest relative to the virtually infinite number of sentence permutations associated with natural language sentence structures, specific templates can often be identified quickly and efficiently. Fixed language components in templates. Second, since the sentence positions of the variable language components in the message template can automatically identify the discrete set of possible replies, the variable language components can also be efficiently identified. For example, taking the above-mentioned " Happy $FESTIVAL" message template as an example, the fixed language component " Happy " can be used as a signal to make the processor 110 know that the subsequent voice input received by the microphone 190 will be the variable language component "$FESTIVAL ".

Although the above-described embodiment of the present invention is illustrated in terms of a wireless telephone 100, the methods and apparatus of the present invention may also include other electronic devices that can provide text messaging, such as a digital personal assistant (PDA).

Thus, the present invention simplifies the steps required to provide and send text messages from a portable electronic device. Text messages may be provided by voice input rather than typing characters from a keypad. Furthermore, the present invention may include open-vocabulary speech recognition to avoid the high memory requirements of the prior art closed-vocabulary speech recognition. Open-vocabulary speech recognition employs speaker-independent sub-word acoustic models designed to cover all occurrences or phonemes of a language. Thus, the user is not limited to using only the pre-set vocabulary, but can also edit the variable language components as described above to include words that cannot be found in dictionaries, such as names of people and places. Thus the text messages provided by the present invention can be highly personalized.

The foregoing detailed description provides preferred exemplary embodiments only, and is not to be considered as limiting the scope, applicability or configuration of the invention. Rather, the detailed description of the preferred exemplary embodiment provides those skilled in the art with an enabling description for implementing the preferred exemplary embodiment of the invention. It being understood that various changes may be made in the function and arrangement of components without departing from the spirit and scope of the invention as claimed in the appended claims.

Claims (9)

1. A method for providing text messages, said method comprising the steps of: receiving speech at an input to an electronic device; Speech recognition is performed on said utterance guided by a user-configured message template stored in a memory associated with said electronic device, wherein said utterance is performed on one of said utterance and said message template Matching means defining speech recognition to build a matching template; and Provide a text message via said matching template, Wherein, at least one of the message templates contains both fixed and variable language components. 2. The method of claim 1, wherein the text message is an SMS message. 3. The method of claim 1, further comprising the step of editing the message template set by the user by receiving characters typed from a keyboard of the electronic device. 4. The method of claim 1, wherein the portion of the text message is a transcription of the utterance. 5. The method of claim 1, wherein the text messages are all transcriptions of the utterances. 6. An electronic device for providing text messages, said device comprising: a microphone for receiving speech; non-volatile memory for storing message templates; and a processor, configured to perform speech recognition of the utterance guided by the message template, the processor matches the utterance with one of the message templates to establish a matching template, and provides text message, Wherein at least one of the message templates contains both fixed and variable language components. 7. The device of claim 6, wherein the text message is an SMS message. 8. The device of claim 6, further comprising a keyboard for editing said message template. 9. The device of claim 6, wherein said processor matches said utterance to a plurality of said message templates and calculates a likelihood value for each of said message templates.

Images