Patent Application
20150142439
The application pertains to systems and methods for providing secure voice control of wireless communications devices. More particularly, the application pertains to such systems and methods which provide authentication of a speaker using multiple identifying indicia.
There is increasing use of “apps” in mobile devices, e.g. tablet computers, smart phones and personal digital assistant (PDA's) to control various building and home automation systems over local area and wide area networks. In addition, there are applications that run on these mobile devices which recognize human speech and perform some task on the device itself or at a central location. In order to improve the human-machine-interface in an automation system, a speech recognition application running on a mobile device which converts speech into digital form and then to other communication protocols suitable for transport on a LAN/WAN, provides a reliable, hands-free, convenient method of use. The '424 application, incorporated herein by reference discloses one such system.
While useful, speech recognition systems can exhibit limitations from a security point of view since speech, not “voice” is being recognized. Speech recognition is much simpler to perform than individual voice recognition. The recognition process however does not necessarily provide a desired level of authentication. Speech recognition is not necessarily tied to an individual. Hence, it would be useful to authenticate the user or speaker in such systems.
While disclosed embodiments can take many different forms, specific embodiments thereof are shown in the drawings and will be described herein in detail with the understanding that the present disclosure is to be considered as an exemplification of the principles thereof as well as the best mode of practicing same, and is not intended to limit the application or claims to the specific embodiment illustrated.
In one aspect, authentication can be implemented prior to speech recognition to provide an increased level of security. In this regard, and to reduce the complexity of voice recognition, it is preferred to target a particular speaker's voice rather than search an extensive database having information associated with a plurality of speakers to find a particular voice.
Advantageously, the particular speaker can be associated with a particular, wireless communication device, for example using a unique smart-phone ID to reduce the complexity of the voice recognition, authentication, process. Another benefit of linking a particular voice and particular device is that certain specific profiles and activities can be authorized subsequent to authentication. For example, a message from a home-owner's phone might produce a different result than a message from a child's or a nanny's mobile phone.
When authenticating a speaker by carrying out a voice recognition activity via a central remote computing station, the wireless device identifier, such as the mobile equipment identifier (MEID), mobile identification number (MIN) or international mobile equipment identifier (IMEI) provide additional originating information so that the voice recognition algorithm can target a specific user. As a result, faster, more reliable and more secure processing can be provided. Additionally authorization can be provided relative to profiles available to a phone/user.
In one aspect, a previously downloaded application being executed on the smart phone digitizes the speech of the individual and sends the information with the mobile device's globally unique identifier to a central computing location. The unique phone ID can be used to identify a particular individual. The authentication process, the voice recognition processing, can use the phone ID as a vector or index into a voice recognition data base which can provide reliable, quicker and secure results.
In another aspect, the wireless communications device can include authentication information for the expected user of that device. In this embodiment, the authentication, and authorization, processing can take place locally at the device. For example, a smart phone. Then the requesting message or command can be transmitted.
The application being executed can include a learning phase to improve security by storing certain phrases from certain speakers and storing the voice patterns with the phone identifier, for example an IMEI.
In one embodiment, an application executing on a mobile phone could transmit a command in the form of digitized speech to a displaced computing facility which, after authentification, would then recognize the command or word, for example “disarm”, from a certain user. The facility could then send the necessary digital data over a network to disarm a specific security system, enable specific lighting scenes, unlock certain doors etc. A small business owner might say “disarm home” to control her home system, or “arm work” to address a change in her business' system.
In an embodiment of a local system, a unique identifier provided by a near field communication (NFC) chip in a smart phone could be used to identify the speaker. Locally stored authentication software can process the incoming audio from the speaker. Once the authentication process has been successfully concluded, and speech recognition carried out, the subsequently recognized command or request can be transmitted to a security system, or any other type of system, for execution.
An exemplary wireless communications device, such as a smart phone, 14 can include a previously downloaded application, app, The appl facilitates authentication and authorization. A user of the phone 14 can verbally speak a command or request into phone 14.
The incoming audio message is digitized and transmitted, using the app executing on the phone 14, along with a phone identifier ID, via a wireless medium to a displaced computing facility 16. The facility 16 could include a programmable processor, along with executable control software to receive and process the digitized voice stream and ID from the phone 14. The facility 16 also includes a voice authentication, recognition, data base 18.
Data base 18 can include voice recognition information for a plurality of individuals. The recognition information for each individual is linked to an individual specific identifier associated with a communications device such as a smart phone, personal digital assistant, computer, tablet or the like, without limitation. For example, the identifier of the phone 14 can be stored in the data base 18 linked to information as to the listed operator of the phone 14. The phone identifier can be used as an index or vector to obtain the pre-stored voice based authentication information from the data base 18 for the specific person associated with the device 14.
The facility 16 can then implement an authentication process with respect to the received, digitized voice sample from phone 14. If the voice is authentified, then the facility 16 can recognize the command or request in the speech steam from the user.
The function, command or other request can then be directed back to system 12 for implementation. For example, system 12 can be disarmed, specific lighting scenes can be enabled, doors can be locked or unlocked, status of areas in the region 12 or environmental conditions can be requested by facility 16 from system 12, all without limitation. Confirmation can be subsequently provided to the phone 14 by the system 12.
In accordance with a method as illustrated in
The incoming digitized audio from phone 14 is processed, as described above in facility 16, using data base 18. If the voice is authenticated and is then authorized, the resultant directive, function or request is forwarded to the system 12 for execution via a WAN, link D. Once system 12 has implemented the order, request or the like, a confirmatory message is forwarded to phone 14 and the user via WAN, link E.
Advantageously, in the combination 10, security is enhanced and over-all processing time can be reduced since the facility 16, upon receipt of the data stream from phone 14, can determine whether the ID of the phone 14 and the associated data stream match the pre-stored voice of the listed operator of the phone 14 without having to retrieve and process extensive quantities of voice information for a large number of individuals from data base 18.
In accordance with a method as illustrated in
Advantageously, in the combination 10-1, security is enhanced and over-all processing time can be reduced since the system 12-1, upon receipt of the data stream from phone 14-1, can determine whether the ID of the phone 14-1 and the associated data stream match the pre-stored voice of the listed operator of the phone 14-1 without having to retrieve and process extensive quantities of voice information for a large number of individuals which might be stored in system 12-1.
An exemplary wireless communications device, such as a smart phone, 14-2 can include a previously downloaded application, app, The appl facilitates authentication and authorization.
A user of the phone 14 can verbally speak a command or request into phone 14-2. The application executed on the phone 14-2 carries out an authentication function, relative to the incoming audio from the user. The received audio, when authentified, can also be processed in phone 14-2 to recognize which command or request has been spoken.
In one embodiment, where the incoming audio corresponds to the pre-stored voice of the authorized user, or owner, the voice stream and mobile phone ID can be transmitted via WAN, links A, B to the displaced computing facility 16-2. Data can also be transmitted from the phone 14-2, via link C to the system 12-2.
The facility 16-2 can process the digitized incoming audio, and if needed carry out a speech recognition function. The request, action, or command can be transmitted from facility 16-2, via link D to system 12-2 for implementation. When the system 12-2 has carried out the requested function, results can be returned to the phone 14-2 via link E.
Alternately, the NFC chip of the phone 14-2 can be enabled so that phone 14-2 and the system 12-2 can communicate directly. The system 12-2 can then implement the order or request.
Advantageously, in the combination 10-2, security is enhanced and over-all processing time can be reduced since the phone 14-2, can directly determine whether the incoming audio matches the pre-stored voice of the listed operator of the phone 14-2 without having to retrieve and process extensive quantities of voice information for a large number of individuals.
From the foregoing, it will be observed that numerous variations and modifications may be effected without departing from the spirit and scope hereof. It is to be understood that no limitation with respect to the specific apparatus illustrated herein is intended or should be inferred.
It is, of course, intended to cover by the appended claims all such modifications as fall within the scope of the claims. Further, logic flows depicted in the figures do not require the particular order shown, or sequential order, to achieve desirable results. Other steps may be provided, or steps may be eliminated, from the described flows, and other components may be add to, or removed from the described embodiments.
| Number | Date | Country | |
|---|---|---|---|
| 61661424 | Jun 2012 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 13710128 | Dec 2012 | US |
| Child | 14603495 | US |
