TITLE: VOICE TRANSLATION METHOD AND PORTABLE INTERACTIVE DEVICE FOR IMPLEMENTING THIS METHOD

DESCRIPTION Background and summary of the invention

The present invention relates to a voice translation method and to a portable interactive device for implementing this method.

As is known, when two persons speak two different languages and both these persons do not understand said languages, the assistance of a third person who is familiar with both languages is required in order to translate for each of the two speakers, so that they can communicate with each other. This third person, who has the function of translating both languages, must obviously have an in-depth knowledge of the linguistic rules, vocabulary, syntax, grammar, etc. of both languages. As is also known, there have been attempts to develop simple and readily understandable universal languages, such as Esperanto, which could allow easy communication between persons speaking different languages. These universal languages, however, have not been particularly successful.

The object of the present invention is to provide a voice translation method which allows two persons who speak different languages to communicate even though neither one of them can understand the language of the other person. Advantageously, the present method provides an instantaneous voice translation and so a person is able to hear a certain phrase translated into his/her language practically in real time, i.e. as soon as the speaker has said it in a language not known to him/her.

This object is achieved by the present invention in the form of a voice translation method, characterized by the following sequence of steps: a) sending a first voice message in a first source language to first means for

receiving/transmitting sound signals; b) recognition of this voice message by these first means (1 ) and checking for its presence in a phrase glossary translated in an intermediate metalanguage; c) encoding the voice message in this intermediate metalanguage; d) transmitting this voice message encoded in this intermediate metalanguage to second means for receiving/transmitting sound signals; e) checking for the presence of the contents of this encoded voice message received by these second means in the phrase glossary of the intermediate metalanguage; f) emitting, from these second means for receiving/transmitting sound signals, a second voice message corresponding to this first voice message translated from this first source language into a second target language.

A further object of the present invention is to provide a portable interactive device for implementing this method, characterized in that it comprises an outer housing which contains a processor connected to a communication module and able to manage voice recognition means, voice synthesizing means and intelligent translation means; these said means are able to recognize a voice message originally sent in a first language and received from these communication means in an intermediate metalanguage and convert this voice message from this intermediate metalanguage into a voice message translated into a second target language.

Brief description of the drawings

Further characteristic features and advantages of the present invention will be better understood from the following description, provided by way of a non-limiting example with reference to the accompanying drawings in which:

- Fig. 1 shows in schematic form two users equipped with a portable interactive device for voice translation according to the present invention; and - Fig. 2 shows a block diagram in which the various steps of the present voice

translation method are indicated.

Description of the preferred embodiment of the invention

With reference to the accompanying drawings and with particular reference to Fig. 1 thereof, 1 denotes the outer housing of a first portable interactive device according to the present invention and comprising a keyboard 2 for accessing menus for configuring and personalizing the device, which are visible via a display 3. This housing 1 is also shown as having a microphone 4 for introducing into this device and sending sound signals such as, for example, voice messages consisting of one or more phrases or one or more words, and a loudspeaker 5 for listening to voice messages sent from another device. By way of example, let us assume that this first device which has the housing 1 is in the possession of a first user U1 , while a second user U2 possesses a second portable interactive device with an outer housing 1 '. The housing 1 and 1 ' of each of these devices is provided internally (see, for example, the partial cross-section of the housing 1 ' of the second device) with a processor 6' powered by an electric battery 7' and provided with a communication module 8' for sending and receiving radio signals. 15' denotes, moreover, a memory module for storing data such as common phrases translated into an intermediate metalanguage or the like. This processor also manages software modules, namely: a voice recognition device 9' of the ASR (automatic speech recognition) type, which has the function of dispensing the user U1 or U2 from having to introduce manually the phrases to be transmitted; a voice synthesizer 10' of the TTS (text-to-speech) type; and an intelligent translator 1 1 ' which performs the translation of the natural language spoken by the user U1 or the user U2 into the intermediate metalanguage.

Considering also the block diagram of Fig. 2, we can see the various steps of the voice translation method according to the present invention performed by each of the devices 1 and 1 ' according to Fig. 1. The user U1 sends a voice message (step f1 ) in a given language: this voice message is shown in Fig. 1 as a sinusoidal signal 12. In

the step f2 the voice recognition device 9 is activated by the processor 6 of the device 1 and a check is performed (step f3) as to whether the phrase contained in the voice message is present in the phrase glossary translated into the intermediate metalanguage and stored in the memory module 15 of the device 1. If it is, this phrase (step f4) is encoded in the intermediate metalanguage and then the first device of the user U1 sends (step f5) a signal 13, relating to the voice message sent and encoded in the intermediate metalanguage, to the second device of the second user U2. If this phrase of the voice message is not present in the intermediate metalanguage, returning to step f3, the context of the phrase is analysed word-by- word (step f6) and the words are translated into the intermediate metalanguage (step f7) by means of the intelligent translator 1 1 ; linguistic rules are introduced (step fδ); and said words of the phrase are encoded in said intermediate metalanguage (step f9), with subsequent sending of the encoded voice message (see step f5 again). This encoded signal 13 is received (step f 15) by the second device of the user U2. At this point the processor 6' checks whether the voice message is a phrase contained in the intermediate language or if it is a set of words, and hence a phrase to be reconstructed (step f10). If this is so, i.e. a phrase or even a single word is recognized, the voice message is translated from the metalanguage into the natural language of the user U2 and this second device emits a sound signal (step f1 1 ), represented by a sinusoidal wave 14, in this language spoken by the user U2 who is therefore able to understand exactly and instantaneously the phrase spoken, in a language unknown to him/her, by the user U1. Returning to step f10, if the voice message received by the second device 1 ' is a phrase to be reconstructed, the following operations are performed: checking the sequence of words of the phrase (step f12); applying a language conversion algorithm (step f13); and generating the phrase (step f14). Therefore, as seen above, once the phrase of the voice message has been reconstructed, the user U2 is able to listen to this voice message translated into his/her language.

Advantageously, therefore, by means of the present voice translation method and the

device for implementing this method, a first person speaks in his/her language by means of a first device 1 described which interprets and converts a spoken phrase into the intermediate metalanguage; this phrase, encoded in the intermediate language, is sent to a second device 1 ' described, in the possession of a second person; this second device of the second person converts this phrase from the intermediate language into the language of this second person. In general, each of these devices possessed by a plurality of users who speak different languages knows exclusively two languages, namely that spoken by the user possessing said device and the intermediate metalanguage, which is known by all the devices. These devices are therefore able to communicate with each other whatever the language spoken by their user.