GB2256959A - Speech recognition device - Google Patents

Speech recognition device Download PDF

Info

Publication number
GB2256959A
GB2256959A GB9113466A GB9113466A GB2256959A GB 2256959 A GB2256959 A GB 2256959A GB 9113466 A GB9113466 A GB 9113466A GB 9113466 A GB9113466 A GB 9113466A GB 2256959 A GB2256959 A GB 2256959A
Authority
GB
Grant status
Application
Patent type
Prior art keywords
speech
words
device
phrases
person
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
GB9113466A
Other versions
GB9113466D0 (en )
Inventor
Glyn * Wallace Nigel
Original Assignee
Glyn * Wallace Nigel
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date

Links

Classifications

    • GPHYSICS
    • G09EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09BEDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B21/00Teaching, or communicating with, the blind, deaf or mute
    • G09B21/009Teaching or communicating with deaf persons
    • GPHYSICS
    • G09EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09BEDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B21/00Teaching, or communicating with, the blind, deaf or mute
    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00Speech recognition
    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L21/00Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
    • G10L21/04Time compression or expansion
    • G10L21/057Time compression or expansion for improving intelligibility
    • G10L2021/0575Aids for the handicapped in speaking

Abstract

An electronic device contained in a conveniently small box suitable for mounting on a wheelchair converts the inarticulate sounds a severely physically disabled and speed impaired person is able to make into normal intelligent conversation. It does this by allowing prerecorded words and phrases to be selected from a display screen by a code made up from a few simple sounds spoken into a microphone. The words and phrases are spoken through a loudspeaker also mounted on the wheelchair. <IMAGE>

Description

PATENT SPECIFICATION SPEECH RECOGNITION DEVICE Device for assisting persons who can only make grunting noises to produce normal speech.

I, Nigel Glyn Wallace, a British Subject of 33 West Hill Road, Foxton, Cambridgeshire CB2 <RTI>6SZ,</RTI> do hereby declare the invention, for which I pray that a patent may be granted me, and the method by which it is to be performed, to be particularly described in and by the following statement : This invention relates to a speaking device for physically disabled people of normal intelligence and hearing capability who are unable to enunciate words but may only emit a limited range of grunting sounds.

Speaking devices for such people are available by which keys or buttons are pressed to actuate speech producers, or text is entered by them from a typewriter type keyboard which is converted into synthesised speech. Many such people however are also highly spastic and severely physically disabled, such that they are unable to operate keys and buttons.

The object of this invention is to enable those who are capable of a limited range of sounds to convert them into intelligible speech.

According to this invention there is provided a combinaton of existing electronic devices which are controlled by a novel form of computer program instructions to enable the complete process to be performed without the aid of a normal able-bodied person or helper. A microphone is placed near to the disabled person's mouth or throat and a loudspeaker is placed at a suitable distance. The devices in between are contained in a convenient box which could be carried on a wheelchair and would be battery operated. The microphone signals are intercepted by a speech pattern recognition circuit which compares the incoming sounds with sound templates held in a Random Access Memory unit. The recognised sounds are used to select from a library of words and phrases held in the same memory unit.These coded words are turned into audio signals by a voice digitising circuit, the words and phrases having already been entered into the memory by a person capable of speech. Alternatively, a synthetic voice can be used with very much less memory requirement from a standard speech synthesiser integrated circuit.

The actions so far described are controlled by a microprocessor circuit which itself responds to grunt sounds. To enable the disabled person to monitor the operation and select the phrases required a standard liquid crystal screen is provided.

Such a device, in accordance with this invention, will now be described, by way of example only, with references to the accompanying diagram of the Grunt Converter.

The Grunt Converter is a device to convert the limited sounds available to a speech impaired person, such as a Dysarthria sufferer, into easily recognised speech. It consists of a microprocessor and memory unit (P) programmed to accept signals from a microphone (M) by way of a standard speech recogniser circuit (R), a program to store these signals in ASCII form in memory, a look-up table to find the words and phrases in the desired order displayed on a screen (S) and a standard speech digitiser (D) to transmit already recorded words and phrases to an amplifier and loudspeaker (L).

The Dysarthria sufferer/user needs to train the speech recogniser to recognise his/her grunts (this is done with a speech therapist in attendance). These sounds are described and entered on a chart on the display screen (S). Another person, male or female, selects the words and phrases most used by the disabled person and enters them into the memory of the Grunt Converter. Both these operations use the standard procedures prescribed by the makers of the units and are carried out using a standard PC computer (C).

They are transferred to the Grunt Converter memory and held by battery (B) back-up permanently in the unit. The conversion programs are held on EPROM permanently and only limited controls are required to operate the unit by the Dysarthria sufferer. The phrases and words are accessed by up to 4 different grunt sounds.

The user can by himself make up the phrases he or she needs from a large vocabulary of words and the phrases can be stored for future use. In cases where the user has difficulty recognising written words the graphic symbols representing words and phrases are displayed on the screen.

The Grunt Converter is contained in a small box with a window display similar to a PC portable (lap top) computer and a limited number of operating buttons (K). It is connected to a microphone (M) for grunt input and a loudspeaker (L) for speech output, and is designed to fit conveniently onto a wheelchair. It is battery operated with a mains adaptor for charging.

A further development of the Grunt Converter uses the principle of Context Selection such that screens of words and phrases are presented such that the words (or phrases) are those which naturally follow the previously selected word. By this means the number of grunts for selection is much reduced and at the same time the range of words available is greatly increased.

Another application for the Grunt Converter configuration is by way of a program for speech training. A speech therapist provides a series of spoken words of increasing difficulty which the speech impaired person tries to match. At each try the device assesses his/her accuracy and the speaking voice, pre-recorded by the therapist, encourages further attempts.

The described grunt converter device has the following advantages : 1. Once set up it can be initiated and closed down by the disabled person without outside help.

2. It enables speech impaired persons to call up intelligible words and phrases without having to use physical movements.

3. It can be used by several such disabled persons together in a group whereby the individuals concerned can select the appropriate speaking voices for themselves.

4. If the speech recognition templates are entered by a speech therapist the device can be used to train a speech impaired person to improve his or her ennunciation.

The apparatus which has been described utilises electronic semiconducting devices of the type that are normally utilised in micro-computers.

Claims (6)

WHAT I CLAIM IS
1. A speech recognition device which allows speech impaired persons to select intelligible words and phrases of their choice by uttering inarticulate sounds without any physical action on their part, using already existing circuits in a novel combination controlled by microprocessor machine instructions which themselves are initiated by the disabled users of the device.
2. A speech recognition device according to claim 1 wherein the words and phrases selected are projected audibly with a pre-recorded speaking voice to the disabled person's choice.
3. A speech recognition device according to claim 1 and 2 wherein the sound activated controls allow several different speech impaired persons to talk to each other with different speaking voices.
4. A speech recognition device according to claims 1 to 3 wherein the speech recognition templates are entered by a therapist for the speech impaired person to attempt to match and thus be trained to speak correctly.
5. A speech recognition device according to claims 1 to 4 wherein means are included to automatically assess and correct the speech impaired person under training.
6. A speech recognition device substantially as hereinfore described with reference to the accompanying diagram.
DESCRIPTION OF DRAWING M Microphone R Pattern Recognition unit P Microprocessor and memory unit S Display Screen D Speech Digitiser unit L Loudspeaker B Battery supply C Setting up computer with keyboard K Small keyboard for switching on by ablebodied person
GB9113466A 1991-06-21 1991-06-21 Speech recognition device Withdrawn GB9113466D0 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
GB9113466A GB9113466D0 (en) 1991-06-21 1991-06-21 Speech recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
GB9113466A GB9113466D0 (en) 1991-06-21 1991-06-21 Speech recognition device

Publications (2)

Publication Number Publication Date
GB9113466D0 GB9113466D0 (en) 1991-08-07
GB2256959A true true GB2256959A (en) 1992-12-23

Family

ID=10697114

Family Applications (1)

Application Number Title Priority Date Filing Date
GB9113466A Withdrawn GB9113466D0 (en) 1991-06-21 1991-06-21 Speech recognition device

Country Status (1)

Country Link
GB (1) GB9113466D0 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1083769A1 (en) * 1999-02-16 2001-03-14 Yugen Kaisha GM &amp; M Speech converting device and method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1083769A1 (en) * 1999-02-16 2001-03-14 Yugen Kaisha GM &amp; M Speech converting device and method
EP1083769A4 (en) * 1999-02-16 2007-05-30 Yugen Kaisha Gm & M Speech converting device and method
US7676372B1 (en) 1999-02-16 2010-03-09 Yugen Kaisha Gm&M Prosthetic hearing device that transforms a detected speech into a speech of a speech form assistive in understanding the semantic meaning in the detected speech

Also Published As

Publication number Publication date Type
GB9113466D0 (en) 1991-08-07 grant

Similar Documents

Publication Publication Date Title
Leather Second-language pronunciation learning and teaching
US4661916A (en) System for method for producing synthetic plural word messages
US5429513A (en) Interactive teaching apparatus and method for teaching graphemes, grapheme names, phonemes, and phonetics
US20040143430A1 (en) Universal processing system and methods for production of outputs accessible by people with disabilities
US6526395B1 (en) Application of personality models and interaction with synthetic characters in a computing system
Gustafson et al. AdApt-a multimodal conversational dialogue system in an apartment domain
Cress et al. Common questions about AAC services in early intervention
Greene et al. Perception of synthetic speech produced automatically by rule: Intelligibility of eight text-to-speech systems
US5774860A (en) Adaptive knowledge base of complex information through interactive voice dialogue
Delgado et al. Spoken, multilingual and multimodal dialogue systems: development and assessment
US6230135B1 (en) Tactile communication apparatus and method
US4984177A (en) Voice language translator
US4757541A (en) Audio visual speech recognition
Lindblom On the notion of possible speech sound
US4641343A (en) Real time speech formant analyzer and display
US20010056345A1 (en) Method and system for speech recognition of the alphabet
US5289521A (en) Audio/telecommunications system to assist in speech and cognitive skills development for the verbally handicapped
US4749353A (en) Talking electronic learning aid for improvement of spelling with operator-controlled word list
Cooper How is language conveyed by speech
Zeng et al. Audio-visual affect recognition through multi-stream fused HMM for HCI
CN101772015A (en) Method for starting up mobile terminal through voice password
Lachs et al. Crossmodal source identification in speech perception
Ramakrishnan et al. Speech emotion recognition approaches in human computer interaction
Tran et al. Improvement to a NAM-captured whisper-to-speech system
US7299188B2 (en) Method and apparatus for providing an interactive language tutor

Legal Events

Date Code Title Description
WAP Application withdrawn, taken to be withdrawn or refused ** after publication under section 16(1)