Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Patent
1997-09-26
2000-10-03
Dorvil, Richemond
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
704247, 704275, G10L 1506, G10L 1522
Patent
active
061285943
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a process of voice recognition and to a system of voice command in a "harsh" environment, that is to say one that is strongly disturbed not only by noise, but by various physical and physiological factors which may act on the user.
2. Discussion of the Background
In a harsh environment such as for example that prevailing in the cockpit of a military aircraft, the apparatuses for voice capture, restitution and processing are not optimized, and are even, mostly, poorly implemented. The radio communications between the crew and other speakers are of poor quality. The users communicate with several other speakers (other aircraft, ground stations, the other members of the crew, their own voice back again). These communications are monophonic, affected by interference, fairly unintelligible and are not hierarchized. Furthermore, various physical and physiological factors may disturb the capabilities of the user. Among these factors may be noted in particular the load factor of the aircraft, the oxygen mask, high respiratory pressure, pilot stress. These poor communications, when added to the environmental noise, contribute significantly to the fatigue of crews, and may even harm their health. The helmets which they wear protect them hardly at all from such noise. The only means available to them to try to make these communications somewhat more intelligible are controls for adjusting sound level, this being far from satisfactory. The various apparatuses which implement these sound communications are dissimilar and their characteristics are not always completely compatible. Furthermore, the growing complexity of on-board systems entails an overload of work for the operator, provoking stress, loss of concentration and eventually of efficiency, this being particularly injurious to the pursuance of a mission under the best conditions. In this context, a voice command device with high recognition rate is particularly beneficial since it makes it possible to control a complex system in a relatively direct and natural way without mobilizing the entire attention of the operator, something which is far from being the case when using tactile surfaces, keyboards, multi-position buttons.
SUMMARY OF THE INVENTION
The subject of the present invention is a voice command system, in particular for a noisy environment, capable of achieving a high recognition rate (typically 95%) with a vocabulary of the order of a few hundred words, and phrases which may contain 10 words and more.
The process of the invention is characterized in that it consists in performing a coarse recognition, then after comparing with stored acoustic references, in supplying the N best phrases recognized, in choosing from these N best phrases, by comparing them with models of probable dialogues and by performing an acoustico-phonetic decoding, the most likely phrase.
The voice recognition system in accordance with the invention includes an acquisition device linked to an acoustico-phonetic decoding device, itself linked to a recognition supervision device, as well as to a voice recognition device, which is itself linked to a dictionary, to the recognition supervisor and to a syntax recognition device, the latter being linked to a dialogue storage device, which is linked to the supervisor.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention will be better understood on reading the detailed description of an embodiment, taken by way of non-limiting example and illustrated by the appended drawing in which: accordance with the invention.
DISCUSSION OF THE PREFERRED EMBODIMENTS
The invention is described below with reference to an aircraft audiophone system, in particular for a combat aircraft, but it is of course not limited to such an application and can be implemented equally well in other types of vehicles (land or sea) and in fixed installations, in particular in very noisy surroundings, such as, for example, in metallurgical plants. The user of this system i
REFERENCES:
patent: 4593403 (1986-06-01), Kishi et al.
patent: 4637045 (1987-01-01), Noso et al.
patent: 4725956 (1988-02-01), Jenkins
patent: 4797924 (1989-01-01), Schnars et al.
patent: 4827518 (1989-05-01), Feustel et al.
patent: 5058081 (1991-10-01), Gulli et al.
patent: 5214707 (1993-05-01), Fujimoto et al.
patent: 5241619 (1993-08-01), Schwartz et al.
patent: 5337251 (1994-08-01), Pastor
patent: 5390278 (1995-02-01), Gupta et al.
patent: 5500903 (1996-03-01), Gulli
patent: 5511009 (1996-04-01), Pastor
patent: 5572623 (1996-11-01), Pastor
patent: 5592385 (1997-01-01), Katayama et al.
patent: 5677990 (1997-10-01), Junqua
patent: 5712957 (1998-01-01), Waibel et al.
patent: 5805772 (1998-09-01), Chou et al.
patent: 5826232 (1998-10-01), Gulli
patent: 5848389 (1998-12-01), Asano et al.
patent: 5937383 (1999-08-01), Ittycheriah et al.
patent: 5987142 (1999-11-01), Courneau et al.
patent: 5991721 (1999-11-01), Asano et al.
Kanno et al., "Generalized Cepstral Modeling of Degraded Speech and Its Application to Speech Enhancement," IEICE Trans. Fundamentals, vol. E76-A, No. 8, Aug. 1993, pp. 1300-1307.
Wrench et al., A New Approach to Noise-Robust LPC, Proceddings ICASSP, IEEE Acoustics, Speech, and Signal Society, vol. 1, Apr. 1987, pp. 305-307.
Gulli Christian
Pastor Dominique
Reynaud Gerard
Dorvil Richemond
Lerner Martin
Sextant Avionique
LandOfFree
Process of voice recognition in a harsh environment, and 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 Process of voice recognition in a harsh environment, and device , we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process of voice recognition in a harsh environment, and device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-204473