Audio control method and audio controlled device

Details Audio control method and audio controlled device Audio control method and audio controlled device

1999-05-19

2004-05-04

Chawan, Vijay (Department: 2654)

Data processing: speech signal processing, linguistics, language

Speech signal processing

Application

C704S270000, C704S235000, C704S260000, C704S272000

Reexamination Certificate

active

06732078

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an audio control method for controlling a data processor with a group of audio commands, in which method information is presented on the display device of the data processor, and at least one control field, to which a predetermined function is assigned, is formed on the display device. The present invention also relates to a device controlled with audio commands, which comprises an audio recognition device, a data processor, a display device for presenting information, means for forming at least one control field on the display device, and means for assigning a predetermined function to said control field.
2. Brief Description of Related Developments
Generally, the purpose of audio control is to facilitate the use of various devices. Such audio control applications include, for example, different devices controlled with speech. Speech control applications are developed, for instance, for computers and telecommunication terminals, such as mobile phones and landline network telephones. With speech control, the user can control a computer by uttering different command words aloud, wherein the user does not have to use the keyboard of the computer for entering these commands. In a speech controlled telecommunication terminal, the user can select the telephone number by saying it aloud, typically one number at a time.
Instructing the computer with voice commands rather than using a pointing device such as a mouse also has significant benefits especially in small, communicator type terminal devices, such as Nokia 9110 Communicator, where the use of the keyboard and/or the pointing device may not be so convenient than the use of bigger keyboard and/or pointing device of e.g. desktop computers.
These speech recognizers are generally based on fixed vocabulary speech recognition or phoneme-based speech recognition. In the fixed vocabulary speech recognition, the device tries to select from a specified vocabulary the word which best corresponds to the word uttered by the user. It is also possible to implement such speech controlled devices in such a way that the user can instruct the device command words with his/her own voice, wherein the device recognizes best the words uttered by that user. The purpose of the speech control methods based on phoneme recognition is to recognize phonemes uttered by the user and to form words on the basis of these phonemes. Such devices based on phoneme recognition are, however, more complex and more expensive than fixed vocabulary speech recognition devices. Furthermore, especially in noisy circumstances the recognition is not as reliable with speech recognition devices based on phonetic recognition as with fixed vocabulary recognizers.
For implementing speech control in devices, in which it is necessary to use only a few command words or numbers, such fixed vocabulary recognizers are well suited. Nevertheless, the aim has been to accomplish speech control also in devices, during the use of which it might be necessary to utter a variety of command words, the command words varying in different situations. For example, when utilizing the Internet data network, it is possible to set up a connection by using several different addresses. Such a browser program for the Internet data network, so-called www browser (world wide web), has recently become common in computers. It is even possible to equip mobile telecommunication devices with such a www browser program, for examining data accessible via the Internet data network. Such a www browser program contains certain standard functions which are largely similar, irrespective of where the data is retrieved from. However, the data retrieved from the Internet data network, for example HTML pages (HyperText Markup Language), may contain active locations, for example links to other Internet addresses, e.g. URL (Uniform Resource Locator), option buttons etc. with varying names, contents and references. In practice, the recognition of these variable names is not possible when using recognition methods and devices of prior art, based on fixed vocabulary speech recognition. On the other hand, especially links can be composed of very long character strings, which the user has to be able to define accurately without misspellings when s/he wants to move to the location indicated by the link. Thus, the speech recognition methods and devices of prior art, based on phonetic recognition are not sufficiently reliable for implementing practicable speech recognition in said browser applications.
Using voice control for www browsing has the difficulty that the links are often long and complex, frequently containing numbers and other non text symbols. This makes them unnatural for a user to say in voice controlled browsing. This problem was solved earlier by instructing the user to say the name of the link (e.g., “Microsoft” for “www.microsoft.com”, or “Nokia” for “www.nokia.com”, . . . ), and then the technique known as speech recognition from text (SRFT) can be used to find the closest match of the input utterance to the currently displayed web links.
SRFT method creates speech recognition models based on text input. From each text entity an acoustic model which represents the spoken equivalent to the text entity is created. The acoustic models are then used to recognize which of the alternative text entities is uttered (if any). SRFT method relies on knowing (or creating) the phonetic structure of the links, making it possible to identify how the user should utter each link name.
Wireless Application Protocol (WAP) is a system architecture specifically designed for use in low bandwidth environment using terminals with varying, often limited, capabilities. Not all terminals are able to display images, for example. A central object of WAP is the WAP gateway (WAP gw), through which all of the traffic between communicating parties (e.g. the terminal and a content server) flows.
The WAP is capable of displaying normal HTML files to the user by converting the HTML to Wireless Markup Language (WML), which is a markup language specifically designed for WAP, in the WAP gateway. Of course the WML can be used independently from the HTML.
Because a small terminal, such as a portable phone, usually cannot display images, it is necessary to offer a textual replacement for an image. This can be done by using the ALT attribute of the image in the HTML, if one exists (e.g. <a href=“main.html”><img border=0 src=“img00253.gif” ALT=“Jack's photo”></a>). The ALT attribute of the IMG tag will be displayed when the pointing device is placed on top of the image containing the link. If an image is used as a link, a text tag, very similar to a voice tag, must be created to be used as the link name if no ALT attribute (or equivalent) directive exists.
When terminals with text and voice I/O are used for www browsing for example in WAP environment, it is impossible for the user to distinguish between different pictures which are used as hyperlinks (i.e. ‘<a href=foo.html><img src=linkpicjpg></a>’type of links), since it is impossible to tell what the picture would tell to the user. Thus, it is very difficult to make a voice tag out of it, and the link name would be either the actual URL the link points to, or something very uninformative like ‘[IMAGE]’. The fact that the name of the image usually does not provide too much information does not make it any easier. Too often the target URL is useless as well, since the target page may be accessed through a common gateway interface (cgi), which can have multiple arguments, or the URL contains multiple random digits and letters, which are difficult to speak and provide no information about the page the link points to. The common gateway interface means computer programs running on a webserver that can be invoked from a www page at the browser.
There is also a possibility that the user of the www browser selects a page, which contains

Affiliated with

Also associated with

Rating

Audio control method and audio controlled device does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Audio control method and audio controlled device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Audio control method and audio controlled device will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3187333