FR2784206A1 - Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized - Google Patents
Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized Download PDFInfo
- Publication number
- FR2784206A1 FR2784206A1 FR9812383A FR9812383A FR2784206A1 FR 2784206 A1 FR2784206 A1 FR 2784206A1 FR 9812383 A FR9812383 A FR 9812383A FR 9812383 A FR9812383 A FR 9812383A FR 2784206 A1 FR2784206 A1 FR 2784206A1
- Authority
- FR
- France
- Prior art keywords
- recognition
- currency
- voice
- converting
- expressed
- 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- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/02—Digital computers in general; Data processing equipment in general manually operated with input through keyboard and computation using a built-in program, e.g. pocket calculators
- G06F15/025—Digital computers in general; Data processing equipment in general manually operated with input through keyboard and computation using a built-in program, e.g. pocket calculators adapted to a specific application
- G06F15/0258—Digital computers in general; Data processing equipment in general manually operated with input through keyboard and computation using a built-in program, e.g. pocket calculators adapted to a specific application for unit conversion
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Abstract
Description
II
CONVERTISSEUR DE DEVISES A COMMANDE VOCALE VOICE-CONTROLLED CURRENCY CONVERTER
L'invention concerne un convertisseur de monnaie utilisant une interface vocale entre le calculateur et l'utilisateur. Différents types de convertisseurs sont déjà utilisés dans l'art antérieur, mais tous ont en commun d'être muni d'un clavier numérique ou alphanumérique pour la saisie des nombres à convertir. Ces dispositifs demandent ainsi un certain nombre de manipulations pour entrer les chiffres manuellement. Ces manipulations peuvent vite s'avérer fastidieuses si l'utilisateur n'a pas une grande liberté de mouvement, que ce soit de manière temporaire ou permanente, comme cela peut être le cas pour The invention relates to a currency converter using a voice interface between the computer and the user. Different types of converters are already used in the prior art, but all have in common that they are provided with a numeric or alphanumeric keyboard for entering the numbers to be converted. These devices thus require a certain number of operations to enter the numbers manually. These manipulations can quickly prove tedious if the user does not have a great freedom of movement, whether temporarily or permanently, as can be the case for
des personnes handicapées.people with disabilities.
Le dispositif objet de l'invention répond à ces inconvénients en proposant un convertisseur utilisant la parole et ne nécessitant que des interventions manuelles restreintes. Ce dispositif est applicable, de manière plus générale, à des calculatrices ou à des opérations The device object of the invention addresses these drawbacks by proposing a converter using speech and requiring only limited manual intervention. This device is applicable, more generally, to calculators or operations
numériques nécessitant peu de fonctions. digital devices requiring few functions.
La gestion des commandes vocales nécessite l'emploi de microphones qui enregistrent ces commandes et, dans le même temps, le bruit environnant. La principale difficulté consiste donc à reconnaître les mots et les chiffres prononcés tout en faisant abstraction de ce bruit parasite. Le dispositif objet de l'invention permet de convertir un montant monétaire exprimé vocalement dans une première devise en une deuxième devise qui est aussi exprimée vocalement. Ce dispositif comporte des moyens de reconnaissance vocale pour l'acquisition du montant dans la première devise, des moyens de conversion numérique, et des moyens de synthèse vocale. Le moyen de reconnaissance vocale comporte une base de reconnaissance contenant uniquement les dix chiffres ainsi qu'un nombre de mots inférieur à vingt. Il comporte aussi un moyen de validation du montant acquis par la reconnaissance vocale utilisant un moyen de mesure de corrélation entre chaque terme prononcé et les termes présents dans la base vocale ainsi qu'un calculateur pour évaluer le taux de reconnaissance globale en considérant le minimum des mesures de corrélation. Cette validation est effectuée lorsque le taux de reconnaissance globale dépasse une valeur de seuil paramétrable et par une The management of voice commands requires the use of microphones which record these commands and, at the same time, the surrounding noise. The main difficulty therefore consists in recognizing the words and numbers spoken while ignoring this parasitic noise. The device which is the subject of the invention makes it possible to convert a monetary amount expressed vocally in a first currency into a second currency which is also expressed vocally. This device includes voice recognition means for acquiring the amount in the first currency, digital conversion means, and voice synthesis means. The voice recognition means comprises a recognition base containing only the ten digits and a number of words less than twenty. It also includes a means of validating the amount acquired by voice recognition using a means of correlation measurement between each term spoken and the terms present in the voice database as well as a calculator to assess the overall recognition rate by considering the minimum of correlation measures. This validation is performed when the overall recognition rate exceeds a configurable threshold value and by a
action manuelle.manual action.
Avantageusement, la base de reconnaissance est Advantageously, the recognition base is
créée par un apprentissage auprès d'un panel de locuteur. created by learning from a panel of speakers.
L'invention concerne aussi un procédé de conversion d'un montant monétaire exprimé vocalement dans une première devise en une deuxième devise exprimée vocalement. On procède à l'acquisition vocale d'un montant exprimé dans la première devise, puis on compare les phonèmes acquis avec les phonèmes de référence enregistrés dans une base de reconnaissance contenant uniquement les The invention also relates to a method for converting a monetary amount expressed by voice in a first currency into a second currency expressed by voice. We proceed with the voice acquisition of an amount expressed in the first currency, then we compare the phonemes acquired with the reference phonemes recorded in a recognition base containing only the
dix chiffres ainsi qu'un nombre de mots inférieur à vingt. ten digits and fewer than twenty words.
Enfin, on valide la reconnaissance lorsque le taux de corrélation est supérieur à une valeur seuil, et on procède Finally, we validate the recognition when the correlation rate is greater than a threshold value, and we proceed
à l'adresse d'un calculateur avec le montant ainsi reconnu. to the address of a calculator with the amount thus recognized.
La figure 1 montre le schéma de fonctionnement Figure 1 shows the operating diagram
du dispositif.of the device.
La figure 2 donne un exemple de carte Figure 2 gives an example of a map
permettant d'implémenter ce dispositif. allowing to implement this device.
Le dispositif est équipé de au moins un bouton (11) permettant d'entrer les données. La conversion se The device is equipped with at least one button (11) for entering the data. The conversion is
déroule en plusieurs étapes.takes place in several stages.
Par un appui continu sur le bouton d'entrée By continuous pressing of the input button
(1), on actionne le système de reconnaissance. (1), the recognition system is activated.
L'utilisateur, tout en maintenant sa pression sur ce bouton, prononce les chiffres de la partie entière du nombre un à un (2). La séparation des termes peut se faire de deux manières. La première consiste à relâcher le bouton entre chaque terme et la seconde, utilisée préférentiellement, consiste à repérer les temps morts entre les termes pour permettre un processus de reconnaissance agréable pour l'utilisateur. Ledit utilisateur prononce ensuite la monnaie d'origine et, enfin, prononce, si cela est nécessaire, les deux chiffres de partie décimale. Après cette opération, il relâche ce bouton et le système essaie de reconnaître les données entrées (3) Le dispositif contient une base de reconnaissance dans laquelle sont mémorisés les différents mots et chiffres, avec plusieurs prononciations possibles pour ces termes. La reconnaissance consiste alors à rechercher le terme de la base qui approche le mieux le terme prononcé. Cette approximation fournit ensuite une valeur mesurant le taux de reconnaissance, qui sera The user, while holding down this button, pronounces the digits of the whole part of the number one by one (2). There are two ways to separate terms. The first consists in releasing the button between each term and the second, used preferentially, consists in identifying the dead times between the terms to allow a process of recognition pleasant for the user. Said user then pronounces the original currency and, finally, pronounces, if necessary, the two decimal digits. After this operation, he releases this button and the system tries to recognize the entered data (3) The device contains a recognition base in which the different words and numbers are stored, with several possible pronunciations for these terms. Recognition then consists in seeking the term of the base which best approaches the term pronounced. This approximation then provides a value measuring the recognition rate, which will be
avantageusement un pourcentage, pour chaque terme reconnu. advantageously a percentage, for each recognized term.
Le pourcentage le plus grand donne ainsi le terme le mieux The larger percentage gives the better term
reconnu.recognized.
Le dispositif considère alors la procédure comme valide à plusieurs conditions. Les conditions de forme sont qu'un mot constituant une devise soit prononcé et que le nombre ait un format correct, c'est-à- dire que l'on obtienne un nombre quelconque, suivi de la mention de devise et, éventuellement, un nombre à deux chiffres. La dernière condition de validité consiste à ce que le minimum des pourcentages de reconnaissance soit supérieur à un seuil donné. La valeur de ce seuil est paramétrable dans le The device then considers the procedure to be valid under several conditions. The formal conditions are that a word constituting a currency is pronounced and that the number has a correct format, that is to say that one obtains any number, followed by the mention of currency and, possibly two-digit number. The last condition of validity is that the minimum of the recognition percentages is higher than a given threshold. The value of this threshold can be configured in the
dispositif.device.
Si le nombre pas valide, un mot ou un son est émis pour inviter à recommencer la saisie (7). Sinon, la synthèse prononce le nombre en clair, avec la partie entière, la devise, et la partie décimale si elle est non If the number is not valid, a word or sound is emitted to invite you to start typing again (7). Otherwise, the summary pronounces the number in clear, with the whole part, the currency, and the decimal part if it is not
nulle.nothing.
Dans une première variante, la conversion peut s'effectuer automatiquement après la fin de la In a first variant, the conversion can be carried out automatically after the end of the
prononciation par le moyen de synthèse vocale (5 et 6). pronunciation by means of speech synthesis (5 and 6).
Dans une seconde variante avantageuse, la conversion peut s'effectuer après validation par l'utilisateur de la pertinence de la reconnaissance. Cette validation peut être faite par l'appui d'un deuxième bouton ou par la prononciation d'un mot de confirmation dont la reconnaissance n'est elle-même pas validée pour éviter des processus récursifs. Dans les deux cas, chacune des validations ultérieures permet de réentendre la somme In a second advantageous variant, the conversion can be carried out after validation by the user of the relevance of the recognition. This validation can be done by pressing a second button or by pronouncing a confirmation word whose recognition is not itself validated to avoid recursive processes. In both cases, each of the subsequent validations makes it possible to hear the sum again
entrée et la somme convertie.entry and the sum converted.
Avantageusement, l'appareil s'éteint tout seul au bout d'un temps fixe si aucune action n'a été effectuée Advantageously, the device switches off by itself after a fixed time if no action has been taken
entre-temps (8).meanwhile (8).
La carte permettant de réaliser les fonctions précédentes peut, de manière préférentielle mais non obligatoire, comporter les éléments distincts suivants: un bouton (11) permettant de mettre en marche le dispositif et un microprocesseur (12) traitant les différentes informations. Elle contient en outre des moyens d'alimentation, comme une pile (19) et un régulateur de tension (18) ainsi qu'un oscillateur (17) qui effectue une synchronisation temporelle. La partie effectue les fonctions vocales peut se composer d'un convertisseur audionumérique (13), d'un microphone (14), ainsi que d'un amplificateur (16) et d'un haut-parleur (15). Cette disposition décrit des éléments séparés les uns des autres, mais l'homme du métier peut aisément trouver des composants The card making it possible to carry out the above functions may, preferably but not necessarily, include the following separate elements: a button (11) for switching on the device and a microprocessor (12) processing the different information. It also contains supply means, such as a battery (19) and a voltage regulator (18) as well as an oscillator (17) which performs time synchronization. The part performing the voice functions may consist of a digital audio converter (13), a microphone (14), as well as an amplifier (16) and a speaker (15). This arrangement describes elements that are separate from each other, but those skilled in the art can easily find components.
effectuant plusieurs fonctions.performing several functions.
Claims (3)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR9812383A FR2784206A1 (en) | 1998-10-02 | 1998-10-02 | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized |
FR9901567A FR2784207A1 (en) | 1998-10-02 | 1999-02-10 | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR9812383A FR2784206A1 (en) | 1998-10-02 | 1998-10-02 | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized |
Publications (1)
Publication Number | Publication Date |
---|---|
FR2784206A1 true FR2784206A1 (en) | 2000-04-07 |
Family
ID=9531146
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
FR9812383A Withdrawn FR2784206A1 (en) | 1998-10-02 | 1998-10-02 | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized |
Country Status (1)
Country | Link |
---|---|
FR (1) | FR2784206A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2811448A1 (en) * | 2000-07-07 | 2002-01-11 | Paul Galbois | Device for converting price in one currency to another has microphone to receive value for conversion and loudspeaker for reproducing voice synthesized result of operation |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06175689A (en) * | 1992-12-07 | 1994-06-24 | Ricoh Co Ltd | Voice recognition reaction device |
EP0844569A1 (en) * | 1996-11-22 | 1998-05-27 | Caisse Régionale de Crédit Agricole Mutuel du Gard | Speech recognition device for currency conversion |
-
1998
- 1998-10-02 FR FR9812383A patent/FR2784206A1/en not_active Withdrawn
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06175689A (en) * | 1992-12-07 | 1994-06-24 | Ricoh Co Ltd | Voice recognition reaction device |
EP0844569A1 (en) * | 1996-11-22 | 1998-05-27 | Caisse Régionale de Crédit Agricole Mutuel du Gard | Speech recognition device for currency conversion |
Non-Patent Citations (2)
Title |
---|
PATENT ABSTRACTS OF JAPAN vol. 018, no. 513 (P - 1805) 27 September 1994 (1994-09-27) * |
TOHKURA: "A weighted cepstral distance measure for speech recognition", IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 86), vol. 1, 7 April 1986 (1986-04-07) - 11 April 1986 (1986-04-11), TOKYO, JP, pages 761 - 764, XP002088868 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2811448A1 (en) * | 2000-07-07 | 2002-01-11 | Paul Galbois | Device for converting price in one currency to another has microphone to receive value for conversion and loudspeaker for reproducing voice synthesized result of operation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP0974221B1 (en) | Radiotelephone voice control device, in particular for use in a motor vehicle | |
EP1585110B1 (en) | System for speech controlled applications | |
US20140316762A1 (en) | Mobile Speech-to-Speech Interpretation System | |
KR101183340B1 (en) | Efficient multimodal method to provide input to a computing device | |
CA2925930A1 (en) | Method for dialogue between a machine, such as a humanoid robot, and a human interlocutor; computer program product; and humanoid robot for implementing such a method | |
US20100178956A1 (en) | Method and apparatus for mobile voice recognition training | |
EP1769489B1 (en) | Voice recognition method and system adapted to non-native speakers' characteristics | |
FR2883095A1 (en) | DISTRIBUTED LANGUAGE PROCESSING SYSTEM AND METHOD OF TRANSMITTING INTERMEDIATE SIGNAL OF THIS SYSTEM | |
FR3058291B3 (en) | ACCESSIBLE ELECTRONIC DOOR | |
JP2004037721A (en) | System and program for voice response and storage medium therefor | |
WO2009071795A1 (en) | Automatic simultaneous interpretation system | |
FR2784206A1 (en) | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized | |
FR2738382A1 (en) | VOICE DIALOGUE SYSTEM FOR AUTOMATED INFORMATION PROVIDING | |
FR2784207A1 (en) | Speech recognition type device for converting between two different currencies has a means for validating spoken input data and alerting the user if his spoken input has been recognized | |
EP1285435B1 (en) | Syntactic and semantic analysis of voice commands | |
JP4230142B2 (en) | Hybrid oriental character recognition technology using keypad / speech in adverse environment | |
CN113113040B (en) | Audio processing method and device, terminal and storage medium | |
WO2006042943A1 (en) | Voice recognition method comprising a temporal marker insertion step and corresponding system | |
FR3058253B1 (en) | METHOD FOR PROCESSING AUDIO DATA FROM A VOICE EXCHANGE, SYSTEM AND CORRESPONDING COMPUTER PROGRAM. | |
US20080133240A1 (en) | Spoken dialog system, terminal device, speech information management device and recording medium with program recorded thereon | |
FR2468161A1 (en) | ELECTRONIC APPARATUS WITH HEARING OUTPUT | |
CA2654961C (en) | Corrector, computer program and method for semantic, syntax and lexical correction of an erroneous expression in a numeric text | |
JPH10198393A (en) | Conversation recording device | |
WO2021239280A1 (en) | System for identifying a speaker | |
FR2867583A1 (en) | Semantic, syntax and lexical electronic proof reader for e.g. dyslexic person, has vocal interaction module to select expression matching most phonetically with dictated expression automatically and replace wrong expression in digital text |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ST | Notification of lapse |