EP2619754A1

EP2619754A1 - Built-in verbal translator having built-in speaker recognition

Info

Publication number: EP2619754A1
Application number: EP11757357.6A
Authority: EP
Inventors: Joël Pedre
Original assignee: Individual
Current assignee: Individual
Priority date: 2010-09-21
Filing date: 2011-08-09
Publication date: 2013-07-31
Also published as: WO2012038612A1; FR2965136A1; US20150039288A1; FR2965136B1

Abstract

The invention relates to a portable electronic translator (1) forming a headset. Said translator (1) comprises at least: a sound capture device (5) with at least one lip microphone (8) and at least one speaking microphone (9). Said capture device (5) is coupled to electronic means (11) so as to determine a current phase of conversation and to automatically adapt the functionalities thereof to said phase.

Description

Verbal translator integrated with perception of integrated interlocutor.

The invention relates to a verbal (or vocal) translator of the portable and autonomous type.

In particular, the invention relates to automatic translation, allowing a first individual speaking in a first language to converse orally with a second individual speaking in a second language, distinct from said first language.

For this purpose, in a device called verbal translator, there is provided a means of translation between the first and second languages, a headset connected to the translation means by a connecting means. This headset is provided with at least one earphone, a microphone and a speaker located on a buccal support bow of said microphone.

This microphone is arranged to capture the words of individuals and then they are transmitted either to the earpiece or to the mouth-arc speaker. Thus, the vis-à-vis the wearer of the translator acoustically perceives the words translated since those pronounced by the wearer, as if these words were emitted directly from the mouth of the wearer. Such a translator is described in document FR2921735A1 or

WO2009080908A1.

Possible improvements have appeared during the secret development of this translator, under the trade name "SpeakWorld ®". These improvements aim on the one hand at the interactivity, the ease and the real effectiveness of the translator's use. On the other hand, it aims at the energy autonomy of such a translator. Indeed, concerning energy autonomy, we know that it is crucial to guarantee a long life of employment and an optimal availability of the translator. It is advisable that at any time, and for any duration of use, this translator is correctly supplied with electrical energy.

This raises various problems. Especially since the power required for the proper functioning of such a translator is not negligible, particularly because of the electrical energy requirements of some of its components, particularly of mouth-horn speakers, which must produce power sound sufficient for perfect hearing by the vis-à-vis the wearer, at a distance and in a possibly noisy environment.

In this context, the batteries and accumulators able to provide such a power often prove to be of an antagonistic size with the lightness and comfort of porting of the translator. In addition, the batteries are made with chemically hazardous products, rare and difficult to recycle, which pushes to limit the integration to a translator.

Also, the charging of the accumulators is often quite impractical, since it forces to have a dedicated charger that increases the size to be transported by the carrier. But also, it is necessary to have a socket connected to the sector. In this case, depending on the country, the different nominal voltages may require to carry, in addition, an additional transformer. While the time required for charging via the sector (e.g. 220V-50Hz), is sometimes long or prohibitive to the use of the translator.

Finally, with the disappearance of fossil fuels, due to the depletion of oil and uranium, the cost of batteries and Access to the sectors, both in terms of purchase and use, is likely to be an obstacle to widespread use of the translator.

This being the case, there are alternative proposals to the questions of supplying electrical energy to portable electronic devices such as a translator. Let's mention some documents related to this question.

US2009120429 discloses a solar powered helmet, with an electronic element and a plurality of photovoltaic solar panels, arranged to be movable between a closed position and an open position.

Document US 61 01 256 describes an integral protective helmet, for example for fireman or motorcyclist. On this headset, to connect its wearer outside a sound point of view, are provided a microphone and a speaker, outside and inside. An amplifier signal coupler couples external audio signals from a microphone upwardly speaking, respectively internal / external or vice versa. A solar power source can be provided.

US2007054705 discloses a non-contact apparatus provided with multiple sources of power supply. A photovoltaic solar panel can be arranged on a high speaker shell of the device in the form of a headset. A removable connector can also be connected to this shell. An optical cable connection is possible. US2005282591 discloses a hand-held mobile phone, with an integrated radio receiver and photovoltaic solar panels at the top of a keyboard and screen case. WO20091 32646 describes the combination of two audio signals from two microphones, to improve the sound reproduction.

Beyond the questions of energy autonomy, let us now mention those of the "SpeakWorld®" type of translators, and their improvements aimed at the interactivity, the ease and the real efficiency of the translator's use.

These improvements aim to make its use even more practical and pleasant, easy and effective. In the field of patents, these improvements are typically described as solutions to various technical problems, revealed by ongoing (secret) research and development.

Regarding in particular the interactivity of use of such a translator, we understand that its ease of use is paramount. Not only to facilitate its handling by a new user. But also so that the user can focus on the conversation in progress, without having to worry or intervene to tell the translator what task is expected of him. In particular, it is understood that during a conversation to translate, several distinct phases intervene, namely: the statement by the wearer in his own language, a speech; the translation of this word into the language of the opposite; - the hearing by the person opposite the wearer of this spoken word in his language; then the statement of another word of response by the vis-à-vis the wearer, in his language not understandable by the wearer; the translation of this word not understandable in the language of the wearer; and finally the hearing by the bearer of this word translated into his language.

In view of the foregoing, it is understood that it would be tedious for the wearer to indicate, for example by actuating touch controls, which phase of the dialogue is found for the translator to adapt his modes of action during of the conversation.

Furthermore, since the translator is in the form of a headset, it is impractical to place such touch controls there. While voice recognition commands may interfere with the conversation.

So, ideally, it would be convenient for the translator to be able to automatically determine what phase of the conversation we are in, and to adapt autonomously (without human instruction) his modes of operation (reception, translation, restitution, etc.) at this stage.

In addition, various practical aspects of using a translator could make it even more attractive. Thus, depending on the situation, the functionalities and therefore the material structures peculiar to such a translator can evolve.

But conversely, in order to avoid burdening and complicating such a translator, a general choice of such hardware is available on the translator, effectively limiting its available features.

In addition, it is common to have one or more electronic devices such as a personal computer, an electronic organizer, a camera or a camera, a media player, etc.

It would therefore sometimes be advantageous to be able to connect such devices with the translator, e.g. to pool their resources (features including information processing, power supply, display, etc.).

Finally, for its transportation and storage, or its presentation for sale or rental, as for its port, it is desirable that the translator can be housed in a receptacle as compact and practical as possible. Include various documents close to these topics. Thus, the document US4949378 discloses a toy in the form of a semi-integral protective helmet with a transparent visor. A microphone on an articulated arm is connected to the toy, and are provided internal and external speakers, the latter located at the top of the hull of the toy, used to emit unintelligible. A noise scrambling selection button is provided.

For the record, recall that the procedure cited the following documents: US20031 15059 which corresponds to WO03052624, US6157727, US2006282269, US2004186727, US2008091407, US2006282269, US20080077387 and

WO2009132646.

The object of the present invention is defined by the claims, to provide an electronic translator, usable using the device according to the invention by following a method determined, to overcome the limitations of translation devices mentioned above by allowing a conversation.

For this purpose, an embodiment of the invention is a portable electronic translator forming a helmet, comprising at least: a sound pickup device arranged on a front branch intended to place this pickup device facing a user's oral position . Said frontal branch being mounted on a main atrium, itself secured to a bow or headrest. The capture device comprising on the one hand at least one mouth microphone arranged towards a rear face of the front branch, and at least one interlocution microphone arranged towards an anterior face of the front branch. A sound reproduction device comprises on the one hand at least one listening speaker integrated in said headset and an interlocution speaker integrated in the front branch to be oriented substantially similar to that of the interlocution microphone . Electronic and logical means are provided in the translator and arranged to ensure the capture, processing, restitution and translation of words.

In one embodiment, the capture device is coupled to the electronic means; at least one interlocution microphone opens in a direction directed towards an interlocutor in a direction forming a conversation axis and is wide front capture fields, while at least one mouth microphone which opens in an opposite direction directed in the direction forming conversation axis is strongly directional rear fields.

The electronic means have means of discrimination of a current phase of conversation, of which one phase of utterance by the wearer which involves a translation in a vis-à-vis language, when a signal from said mouth microphone is greater than another signal from the interlocution microphone. In this case, a speech conversation phase being determined, the electronic means automatically perform a translation processing of said signal from said mouth microphone in a language vis-à-vis.

According to one embodiment, the capturing device is arranged with the microphone of interlocution with wide front capture fields of cardioid type.

According to one embodiment, the capturing device is arranged with the highly directional rear-end microphone, hyper cardioid or barrel type. According to one embodiment, the electronic and logical means are provided at least in part in an atrium of the translator and are arranged to automatically determine phases of: uttered by the wearer in his own language, a word; - translation of this word into the language of the person opposite; hearing by the vis-à-vis the wearer, of this speech translated into his language; utterance of another word of response by the vis-à-vis the wearer, in his language not understandable by the wearer; translation of this word not understandable in the language of the wearer; and hearing by the wearer, of this word translated into his language. According to one embodiment, the translator has at least one photovoltaic sensor

In one embodiment, at least one photovoltaic sensor is on the arch of the translator.

According to one embodiment, the translator has means for viewing the speech / listening state. These means are controlled by the electronic means so that a specific color lighting is activated according to the current conversation phase, another color, clearly distinct visually, being provided for at least one other conversation phase. In one embodiment, the electronic means of the translator have at least one coupling connection to an external electronic device.

According to one embodiment, the electronic means of the translator comprise transcription means which are integrated with display means.

In one embodiment, the translator has a male connector plug, e.g., on a main headset and / or a complementary female plug e.g., on a branch.

Another object of the invention is a translation method, which employs a translator as evoked.

According to the invention, a logic processing performed by the electronic means provides a switching functionality of automatic language with an automatic determination, at each moment, in real time and / or in repetitive intervals, of which is the interlocutor between the translator's wearer and his opposite person who speaks, and the listener. In one embodiment, the electronic means are arranged so that in a listening state, it is expected that the restitution of the signals from the microphone microphone-type pickup device are reduced and those of the other microphone-type pickup device. interlocution are increased, and / or translation processing of which a choice of the language produced and interpreted by the translator and that to be provided via the rendering devices including the headset, are automatically determined.

Various embodiments of the invention are described with reference to the appended figures, in which:

- Figure 1 schematically illustrates a conversation using several translators according to the invention, each of which has means for viewing the speech / listening state; FIG. 2 schematically illustrates a detail of a translator embodiment according to the invention, which has a so-called "unitary" form of means for viewing the speech / listening state, as well as a removable branch at connect by a plug / socket; - Figure 3 schematically illustrates a detail of a translator embodiment according to the invention, coupled to an electronic device and / or network connected by a logical link;

FIG. 4 schematically illustrates a detail of a translator embodiment according to the invention, with another form of detachable branch, and means for adjusting the size of a bow vis-à-vis at least one earpiece where are located means for connecting and assembling said removable branch; and

FIG. 5 schematically illustrates a detail of a translator embodiment according to the invention, with electrical energy supply means of the photovoltaic type, e.g. arranged on a holding bar.

That being said, let us describe non-restrictive examples of realization. In the figures, reference numeral 1 designates an electronic device forming a portable voice translator. A wearer / user 2 puts this translator on his head, here in the manner of a sound headset, when he wants to use it.

Although this is not limiting, the examples show translators 1 of "SpeakWorld ®" type, substantially in accordance with document FR2921 735.

For the use of a voice translator 1, at least one user 2 (or "carrier" or "interlocutor") is needed, one direct and carrying the translator, the other indirect and not necessarily porteu r lu itself a translator 1, but can be equipped.

For example, at a given moment, the interlocutor 2 d irect is called the "main user 2". While the other interlocutor 2, bearer or not, is called "interlocuteu r 2 secondary". Of course, the invention of a SpeakWorld ® (or multiple) vocal translator 1 may also be useful for a group of interlocutors 2 (see Figure 1) of more than two persons wishing to converse. In Figu re 1, we see three users 2 each with a translator 1. A direction 4 is drawn between the mouth positions 3 of the main and secondary interlocutors 2 at a given moment. This direction 4 is called "the axis of conversation" (continuous double dotted line). According to the invention, it has appeared useful and pleasant for the users 2 to improve the sound recording of the verbal (vocal) utterances of the main interlocutor 2.

Two types of sound recordings are mainly known (see: http://fr.wikipedia.Org/wiki/Microphone#La_directivit.C3.A9) in terms of extent: the capture devices 5 with large "capture fields" "(Eg cardioid microphones), and the capture devices 5 with highly directional capture fields (eg hyper cardioid microphones or canon).

It is also known that the translator 1 "SpeakWorld ®" according to document FR2921735 has a front branch 6 on which are placed, close to the conversation axis 4 of the oral position of the interlocutor 2, a device 7 for sound reproduction (loudspeaker) opening in the opposite direction to the interlocutor 2 in the direction 4, and in general at least two devices 5 capturing verbal utterances of the interlocutors 2 main and secondary.

According to the invention, a translator 1 has at least two capture devices 5, eg. on one branch 6. One is oriented outwards that is to say the front of the translator 1, the other inward, that is to say the back of this translator 1. Each inner or rear device 5 serves as a mouth microphone 8. Each external or front device 5 serves as an interlocution microphone 9.

In one embodiment, the mouth microphone 8 forms a device 5 (rear or internal) with highly directional capture fields (e.g. hyper cardioid microphones or barrel). The interlocution microphone 9 forms meanwhile a device 5 (front or outer) with wide capture fields.

For example, this realization is practical for public translations (speech, conference, lecture, for example) where the translation of the words (speeches, speeches, etc.) of the bearer 2 of a translator 1 are preponderant, and the Positional origin of the returns (questions, reactions, etc.) can be of extended origin.

In other embodiments, it is conversely the mouth microphone 8, which is wide field of capture, and the interlocution microphone 9 which is highly directional field of capture.

For example, such a realization is practical for translations or dialogues in a small committee (business negotiations, diplomacy, etc.) where the quality of translation of the words of the one or few vis-à-vis the holder 2 of a translator 1 are relatively preponderant.

The invention therefore makes good use of the fact that naturally, two interlocutors 2 are automatically oriented vis-à-vis when they converse. In fact, their mouth positions 3 (more generally their entire face, including their eyes and mouth together) come opposite or facing each other, to jointly define a conversation axis 4 relatively stable, common and accurate, during a given conversation phase.

Therefore, a highly directional interlocution microphone 9 "targets" or "points" towards the external speaker, which provides a good capture, relatively "concentrated" on the words of this external speaker (to the detriment of background noise surrounding ).

It will be understood that then one of the capture devices 5 has strongly directional (forward) capture fields. When this device 5 opens in the forward direction directed towards the interlocutor

2 in direction 4, the invention therefore makes judicious use of the natural setting compared to the mouth positions

3 combined with a sound recording on a field of limited extent, to somehow "filter" the noise to the conversation.

However, in all cases, the natural setting of the mouth positions 3 places the sound pickup field automatically relatively in front of the mouth 3 vis-à-vis, which ensures optimal capture to the wearer 2. And if this capture is highly directional, the spurious noise surrounding the capture field are only slightly or not recorded, because out of field, which corresponds to a kind of filtering.

Conversely, when the strongly directional capture field is oriented towards the rear, towards the mouth of the carrier 2 of a translator 1, the invention can make use of a reduced sensitivity in terms of sound volume capture by the mouth microphone 8, compared to the superior sensitivity in terms of sound volume pickup by the wide-field 9 speaking microphone. Applications of the invention with a rear-end microphone 8 with a highly directional field plus a front wide-field microphone 9, may be suitable for multi-point dialogues (three or more) in a quiet location, where the microphone is wide-field interlocution 9 facilitates the perception of the words of each vis-à-vis the carrier 2 of a translator.

In still other embodiments of the invention, there are provided several mouth microphones 8 and several interlocution microphones 9, with distinct forms of sound pickup fields.

Thus, in one embodiment of the invention, at least one mouth microphone 8 is wide field of capture and another mouth microphone 8 has a strongly restricted field of capture. In addition, at least one interlocution microphone 9 is wide field of capture and another interlocution microphone 9 is strongly limited capture field.

This is called mixed capture. Electronic and logic means 1 1 of the mixed translator 1 are, in one embodiment, implemented with means 10 for discriminating sound sources, combined with these two microphones 8 back and two microphones 9 before.

For example, the discriminating means 10 evaluates the instantaneous capture quality of each microphone 8-9 and determines which one offers the best rendering of speech. In parallel, the electronic and logical means 1 1 determine the current conversation phase, the "turn of speech", etc. Then, these electronic and logical means 1 1 not only select whether it is appropriate to process the signal from a rear microphone 8 or before 9, but also which (wide or strong field) restricted) offers the optimum sound quality for this conversation phase.

This makes it possible, when such a mixed translator is economically viable, to combine the two approaches and their advantages.

According to the embodiments, at least one highly restricted field capture device 5 is of hyper cardioid or canon technology. While at least one wide capture device 5 is cardioid technology. The invention thus provides a very high sound pickup quality, together with excellent flexibility and positional stability (in proximity and direction by convergence of the two axes 4), which provides comfort and mutual understanding of interlocutors 2 high. In embodiments, the device (s) 5, more precisely the mouth microphone 8 or interlocution 9 have a filter (confused with the means 1 1).

For example, it is a voice learning filter, so that the voice utterances of the two instantaneous interlocutors 2 are picked up and then cleared of noise and / or acoustically focused on the specific sonorities of voice utterances of these two interlocutors 2 snapshots.

In one embodiment the two mouth microphones 8 and interlocution 9 are merged or confused in a single device 5, and leads (eg one before the speaker 2 the other back to the carrier translator 1) also form part of or form source discrimination means 10. For example, the capture devices 5 are microphones produced by Bruel & Kjaer (see http://www.bksv.fr/Products/TransducersConditioning/AcousticTran sducers / Microphones.aspx). Thanks to the invention, it is possible to have a translator 1 at the same time more compact, lightweight and modular.

For this purpose, an embodiment provides that all or most of the electronic and logical means 1 1 of a translator 1 are integrated with a main atrium 12. By electronic and logical means 1 1 denotes the electronic cards and processing components information, sound signal production (including any filters eg sound focus), power supply, connectivity for accessories to the headset 12, external connectivity.

Thus, electronic and logical means 1 1 provide in embodiments of the invention, an external connectivity of the translator 1 to other electronic devices eg PDA, computer, wireless network such as wifi, Bluetooth, 3G, GSM, RFID, etc. . or wired like USB, FireWire, RS232, mono / stereo jack, etc.

These electronic and logical means 1 1 are integrated with the main atrium 12 in the illustrated embodiments. But obviously this is not the case of accessories (possibly removable) such as 5 capture devices (microphones) and 7 playback (loudspeaker) and / or discrimination means 10 (discrimination channels), some of which are remote from this main atrium 12. In the embodiment of Figure 2, we see also a plug 15 of a USB or FireWire connection socket is mounted on the main atrium 12. On the branch 6, which is here removable, is mounted the plug Female 16 complementary (to the male) of this USB or FireWire connection socket.

Of course, the electronic and logical means 11 comprise at least one data storage memory 13 (ROM, RAM, EPROM, etc.) which is arranged in such a way as to be able to memorize the various computer programs or software (hardware management of the software). translator 1, translation, filtering, etc.).

Thus, in one embodiment, the translator 1 has a feature called "automatic language switching".

In synthesis, the sound signals picked up or not (silence) by the device 5 of the mouth microphone type 8, as well as the sound signals picked up or not (silence) by the other device 5 of interlocution microphone type 9, are connected logically (typically within electronic and logical means 1 1 integrated to a main headset 12) to determine at each given moment, the current phase of a conversation between interlocutors 2.

The logical processing performed by these means 1 1 (including a software 14 discrimination) for the purposes of this feature of "automatic language switching", is to automatically determine, at each moment-in real time and / or in repetitive intervals- is the interlocutor 1 (the one who carries the translator 1 in question, or the other translator 1) who speaks and the one who listens, in particular.

Typically, if a significant signal of a verbal utterance is perceived by the device 5 of the mouth microphone type 8 strongly Directional features specific to this state or so-called "speech phase" are put in place.

For example, in the state of speech, the restitution of the signals coming from the other interlocution microphone device 9 are reduced and those of the mouth microphone type device 8, and the translation treatments whose choice of the language perceived by the translator 1 and that to be supplied via the sound reproduction device 7 on the branch 6, are automatically determined. Corresponding filterings are obviously implemented

(means 10 and 1 1 in particular), to promote the mutual understanding of interlocutors 1 using this translator 1.

Conversely, a state or "listening" phase provides for example that the restitution of the signals from a device 5 of the mouth microphone type 8 are reduced and those of another device 5 of the microphone type interlocution 9 are increased.

Then, the translation processing, the choice of the language produced and interpreted by the translator 1 and that to be provided via the device 7 including the headset 12, are automatically determined.

Thus, an embodiment provides in a state of speech that the language perceived or spoken by the bearer 1 is French, and the translated language is English, while the translator 1 passes or is automatically maintained in translation mode. verbal statement "(eg Chinese to German).

While at the perception of a silence (relatively durable and marked in loudness) by the device 5 of the mouth microphone type 8, the translator 1 is switched in phase and "listening state", for example after validation. 'a capture significant by the other d ispositif 5-type interlocution microphone 9, and the logical processing is placed in "translation of external speech" (eg German to Chinese).

In Figure 1 we see that the translator 1 has display means 17 of the speech / listening state.

In this embodiment, on the one hand, these means 17 have a green LED (s) 1 8 on the branch 6 here in the vicinity of the devices 5 and 7, that is to say on the outer face of the device. this branch 6. On the other hand, the means 1 7 have another block of LEDs (s) bright 1 9 of red color on the main atrium 1 2, here in the vicinity of the devices 5 and 7 of this atrium 12.

In FIG. 2 it can be seen that the translator 1 has another so-called "unitary" form of display means 17 of the speech / listening state. In this embodiment, these means 1 7 have a single block of LED (s) luminous 1 8-19 variable color (depending on the frequency and / or supply voltage) fully grouped on the branch 6 and also near the devices 5 and 7 of the outer face. These means 1 7 grouped are pi lot by the means 1 0 so that the (or) pavement (s) of LED (s) 1 8-1 9 illuminated light of a cou their determined by a specified state or phase of the translator 1. Another color, clearly distinct visually, is provided for another state or phase. Here, the keypad (s) 1 8- 1 9 bright illuminates in green for the listening state of the carrier 2, and red color for the state of speech of said port r 2.

The interlocutor 2 in front of a carrier 2 of a translator 1 in speech (red) thus sees that it is preferable that he keep quiet and listen when the wearer speaks 2 Conversely he sees that he can express himself verbally when the translator 1 indicates to him in green that the wearer 2 is in a listening state. This avoids overlapping verbal utterances, facilitates processing and improves conversation quality.

In Figure 3, it can be seen that the translator 1 is coupled to an electronic apparatus e.g. PDA, computer, wireless network such as wifi, BlueTooth, 3G, GSM, RFI D, etc. This apparatus 20 is connected by a logic link 21 which can be wired (for example via a cable 22 on which are coupled two "jack" jacks of diameter 1, 5mm).

We also see that the apparatus 20 has display means 23. This is conventional for PCs, PDAs, Ipod®s /

Iphone®, BlackBerry® and other devices 20 Psion® type, Archos®, Android® in particular.

Due to and via the connection 21, various functionalities can be transferred to these means 20, out of the translator 1, eg of its main headset 12 (which may nevertheless include all or some of the other features outside those specific to the branch 6 ).

In one embodiment, visualization means 17 of translator 1 are integrated with the display means 23 of the apparatus 20.

In one embodiment, transcription means 24 are integrated in the display means 23.

Typically, these means 24 allow the written display of: ung lossaire, interactive lexical propositions (based on voice recognition and / or offering in writing 25 eg choices of quasi-homonyms, to be selected, - dictionaries, transcripts of conversations (for example transferable on computer, eg by email or chat, or to a word processor, operating aids (multilingual user guide) of translator 1, etc.).

An example of transcription means 24 is ViaVoice® or Dragon®.

In one embodiment (FIG. 3), written text reading means 26 are integrated and / or accessible via the apparatus 20. Then, via the connection 21 and / or 22, a reading of written text, produced by by or through the apparatus 20 is transmitted orally to the translator 1 (eg to the main atrium 12). Thus, a written text is read and stated by the translator 1.

Note that in Figure 2, the means 7 have a speaker 37 interlocution. In FIG. 3, the means 7 also have a listening loudspeaker 36.

Then, the electronic and logical means 1 1 (including a translation model 27 in particular (FIG. 3) operate to the attention of the carrier 2 equipped with the translator 1, a voice translation of the written text, in the listening language. of the carrier 2 (eg, selected in advance via a control interface 28 of the translator 1, on the latter and / or on the apparatus 20). Typically, the control interface 28 is tactile (screen, buttons ...), but other control means (eg verbal) are provided in embodiments, especially when the risk of interference with the words to be translated is limited. According to embodiments, the translation module 14 (FIG. 2) comprises electronic components and language management programs such as Nuance (http://nuance.fr), Jibbigo (http://www.jibbigo.com) or like.

The invention allows e.g. via the control interface 28 and / or using the means 10-1 1, easy validation of dialogue languages, clarity of sound, etc..

According to the embodiment of FIG. 4, it can be seen that the translator 1 has a male connection plug 15, here on the main atrium 12. For example, the translator 1 is equipped with a USB plug or the like, allowing the connection of the translator 1 for exchanging data, signals or for charging electricity.

The translator 1 also has a complementary plug 16, here female, into which the plug 15 can be connected. Integrated to this plug 16, a thin rod 29 at the distal end (opposite to the complementary plug 16) of which are arranged devices 5 (Mouth 8) that allow when a translation is not desired, to use the structures of the translator 1 as a simple headset and / or speech.

Such a headset is useful with various types of apparatus 20, once connected via links 21 or 22, eg. to a cell phone, a PDA, a walkman, a computer, or the like. According to the embodiment of Figure 4, we see that the translator 1 has a hoop 30 which supports (or integrates) at least one atrium 12.

This arch 30 is mechanically coupled to at least one atrium 12 by adjusting and immobilizing means 31. Typically, these adjustment and immobilization means 31 have a sliding link 32 with ratchets.

To adjust the relative position (up-down / disassembly: see arrow 33) or disconnect to remove one or two atria 1 2 from the translator 1.

In FIG. 3, it can be seen that the translator 1 has at least one photovoltaic (solar) sensor designated at 34. Here, a photovoltaic sensor 34 is on the hoop 30.

A translation process is implemented by the translator 1.

Indeed, men communicate with each other vocally using a greater number of languages. Most countries have at least one official language of their own which differs from languages spoken in other countries. Even in some countries, men do not communicate with each other using the official language of their country, but rather by using a local language, called "patois" for example.

As the means of transport develop in an important way, it becomes easy to travel from one country to another. Similarly, international trade is an important part of global economic activity. The fields of research and exchange of knowledge are also increasingly international, thus polyglot.

As a result, whether it is in the context of a tourist trip or professional activities, a first individual is likely to have to communicate with a second individual in a language that differs from his mother tongue.

It is thus common to learn one or more foreign languages during his schooling. Unfortunately, it is almost impossible to master and speak all the existing languages.

Claims

REVEN DICATIONS

1 - Translator (1) portable electronic helmet, comprising at least: a sound pickup device (5) arranged on a front branch (6) provided for placing the capture device facing a buccal position (3) of a carrier (2); said front branch being mounted on a main atrium (12), itself secured to a bow (30) or headrest; the capture device (5) comprising on the one hand at least one mouth microphone (8) arranged towards a rear face of the front branch (6), and at least one interlocution microphone (9) arranged towards an anterior face frontal branch; a sound reproduction device (7) comprising, on the one hand, at least one listening loudspeaker (36) integrated in said earpiece and an interlocution loudspeaker (37) integrated in the front branch so as to be oriented substantially similarly to that of the interlocution microphone (9); electronic and logic means (1 1) being provided in the translator and arranged for capturing, processing, rendering and translating words, characterized in that the pick-up device (5) is coupled to the electronic means ( 1 1); at least one interlocution microphone (9) is oriented in a direction of travel towards a speaker in a speech axis direction (4) and has a wide front pick-up field, while at least one mouth microphone (8) ) which is oriented in the opposite direction, is directed along the direction of conversation axis (4) and is strongly directional rear field.

2- Translator (1) according to claim 1, characterized in that the electronic means (1 1) have discrimination means (1 0) of a current phase of conversation, including a carrier utterance phase which involves a translation into a language of vis-à-vis, when a signal from the dud it microphone of mouth (8) is of loudness greater than another signal from the microphone Interlocution (9). 3- translator (1) according to one of claims 1 or 2, characterized in that a speech conversation phase being determined, the electronic means (1 1) automatically perform a translation processing said signal from said microphone mouth (8) in a language vis-à-vis. 4- Translator (1) according to one of claims 1 to 3, characterized in that the capture device (5) is arranged with an interlocution microphone (9) with wide front capture fields, cardioid type.

5-translator (1) according to one of claims 1 to 4, characterized in that the capturing device (5) is arranged with a highly directional rear-field microphone (8) of the hyper cardioid type or cannon.

Translator (1) according to one of Claims 1 to 5, characterized in that the electronic and logic means are provided at least partly in an atrium (12) of the translator and are arranged to automatically determine phases of: the bearer in his own language, a word; translation of this word into the language of the opposite person; - hearing by the person opposite the wearer of this spoken word in his language; utterance of another word of response by the vis-à-vis the wearer, in his language not understandable by the wearer; translation of this word not understandable in the language of the wearer; and - hearing by the wearer, of this word translated into his language.

7- Translator (1) according to one of claims 1 to 6, characterized in that the translator has at least one photovoltaic sensor (34). 8- Translator (1) according to claim 7, characterized in that at least one photovoltaic sensor (34) is on the arch (30) of the translator.

9- Translator (1) according to one of claims 1 to 8, characterized in that the translator has means for viewing (17) the state of speech / listening; these means being controlled by the electronic means (11) so that a specific color lighting is activated according to the current conversation phase, another color clearly distinct visually being provided for at least one other conversation phase.

10- Translator (1) according to one of claims 1 to 9, characterized in that the electronic means (11) of the translator have at least one connection (21; 22) for coupling to an electronic device (20) external. 11- Translator (1) according to one of claims 1 to 10, characterized in that the electronic means (11) of the translator comprise transcription means (24) which are integrated with display means (23).

12- Translator (1) according to one of claims 1 to 11, characterized in that the translator (1) has a plug (15) of male connection, eg on a main atrium (12) and / or a complementary female plug (16) eg on a branch (6).

13- Translation method, employing at least one translator (1) according to one of claims 1 to 12, characterized in that a logic processing performed by the electronic means (11) provides automatic language switching functionality with a determination automatic, at each moment, in real time and / or repetitive intervals, which is the interlocutor between the carrier of the translator (1) and his vis-à-vis who speaks, and the listener.

14- The method of claim 13, characterized in that the electronic means (11) are arranged so that in a state or listening phase, it is expected that the return of the signals from the microphone-type pickup device (5) of the mouth (8) is reduced and the restitution of the other microphone-type pick-up device (5) (9) is increased, and / or translation processes including a choice of the language produced and interpreted by the translator (1) and that to be provided via the rendering device (7) are automatically determined.