US RE41080 E1
The present invention is an item location system which relies upon voice activation and responsiveness to identify location(s) of item(s) sought by a user. The system includes a continuous speech recognition digital signal processor, a programmable microprocessor interfaced therewith, voice input and user feedback mechanisms, including audio and/or video feedback. Preferred embodiments utilize audio feedback to the user. The continuous speech recognition engine utilizes Hidden Markov Models to create real time continuous speech recognition and feedback.
1. An item locator system having both voice activation and voice responsive capabilities for location feedback to locate one or more specific goods in a retail store facility, which comprises said system comprising:
a.) a support structure, for physically supporting said system at one or more locations, and functionally containing or connected to the following components:
b.) a continuous speech recognition digital signal processor (DSP), wherein said continuous speech recognition engineDSP utilizes tokens of raw acoustic signals representing utterances or words and matches these against a set of models and then relies upon likelihood to select a most likely model to decode signals for interpretation;
c.) a programmable microprocessor interfaced with, coupled to said speech recognition DSP; , d.) sufficient programming and circuitry contained within said programmable microprocessor to provide for voice activation and voice recognition andaudio and/or visual response to provide item location to a user wherein item and location data are defined by manager input to said system ;
e.) voice input means connectedcoupled to said speech recognition DSP;
f.)at least one memory storage means connected to, coupled to said programmable microprocessor and/or said DSP, for storage of operational inputs, control inputs, and a voice recognition vocabulary for storage of command match and execute functions; and
g.) at least one user feedback unit and connection from, coupled to said programmable microprocessor to said at least one user feedback unit, said at least one user feedback unit ,adapted to provide audio and/or visual feedback selected from the group consisting of audio feedback, visual feedback and combinations thereof, to asaid user in response to an item location query.
2. The system of
3. The system of
4. The system of
5. The system of
6. The system of
7. The system of
8. The system of
an audio feedback component which includes including audio feedback hardware and a software adapter to audibly respond to recognizable voice input, including wherein said audio feedback component uses digital-to-analog conversion; and an output speaker to provide audio feedback to said user.
9. The system of
10. The system of
11. The system of
12. The system of
13. The system of
14. The system of
15. The system of
16. The system of
17. The system of
18. The system of
19. The system of
20. The system of
21. An item locator system for location feedback to locate one or more specific items in a facility, comprising:
at least one processor operable to receive acoustic signals representing utterances or words, match the acoustic signals against a set of models, and select one of the set of models to decode the acoustic signals, wherein the at least one processor is further operable to detect a trigger word for continued interpretation of the utterances or words;
at least one memory coupled to at least one processor for storage of operational inputs and a voice recognition vocabulary of command match and execute functions; and
at least one user feedback circuit coupled to the at least one processor, wherein the at least one user feedback circuit is operable to provide visual or audio feedback to a user in response to an item location query.
22. A locator system comprising:
a local processor residing in a facility;
a database, coupled for access by the local processor, storing location information for each of a plurality of items in the facility;
a remote processor, coupled to communicate with the local processor, operable to update at least a portion of the database;
a voice-driven interface, coupled to the local processor, for accepting acoustic signals corresponding to a request by a user for the location of at least one of the plurality of items; and
a controller, coupled to the local processor, to provide audio or visual signals to the user regarding the location of the at least one of the plurality of items.
23. The locator system of
24. The locator system of
the facility is a retail store; and
the item is a consumer item.
25. A locator system comprising:
at least one processor for processing acoustic signals corresponding to a request by a user for the location of at least one of a plurality of items in a facility;
a database, coupled to the at least one processor, storing location information for each of the plurality of items; and
at least one memory, coupled to the at least one processor, for storage of an originally active vocabulary and a voice recognition vocabulary of command match and execute functions, wherein the at least one processor is configured to access the at least one memory when processing the acoustic signals, and the at least one processor is configured to access the database to provide information to the user regarding the location in the facility of the at least one of the plurality of items.
26. The locator system of
27. The locator system of
28. A locator system comprising:
a plurality of user locations, each comprising a local processor, located in a facility;
a database, coupled for access by the local processor of each of the plurality of user locations, storing location information for each of a plurality of items in the facility; and
wherein each of the plurality of user locations further comprises:
a voice-driven interface for accepting acoustic utterances or words from a user to request the location of at least one of the plurality of items;
at least one memory with a voice recognition vocabulary for storage of command match and execute functions; and
a controller to provide audio or visual signals to the user regarding the location of the at least one of the plurality of items.
29. The locator system of
30. The locator system of
31. The locator system of
1. Field of the Invention
The present invention relates to voice activated/voice responsive item locators, i.e. item directories which direct a user such as a consumer or shopper, to a specific location to view, treat, retrieve, order, purchase or otherwise use the information obtained in the system. Typically, the present invention could be used at retail stores to locate items to be purchased. Alternatively, it could be used at a production facility or distribution facility having a large number of parts, to locate specific parts for an employee. In other embodiments, it could be used in non-commercial entities, such as a public library to locate a particular book. The locator of the present invention relies upon a specific software module to accomplish voice recognition and response, and includes manager programming for customization, updates and modifications.
2. Information Disclosure Statement
The following prior art patents represent various inventions relating to machines involving speech recognition for voice-based operation and thus illustrate known voice recognition applications:
U.S. Pat. No. 5,111,501 to Masanobu Shimanuki describes a telephone terminal device equipped with a transmitter microphone, a receiver, a speech recognition unit that receives and recognizes speech signals from the transmitter microphone and a circuit to reduce the level of signals send from a telephone network to the receiver when the speech recognition unit receives speech signals from the transmitter microphone. Further, this device is preferably equipped with a speech reproduction unit that reproduces the speech information stored in a memory, in response to the information of recognition result from the speech recognition unit, and a circuit that prevents transmission of signals from the telephone network to the receiver when the regenerated speech information in sent to the receiver. Furthermore, it is desirable for this device to be provided with a circuit that prevents generation of ringing tones when an incoming call arrives.
U.S. Pat. No. 5,136,634 to David C. Rae et al. describes voice operated facsimile machine network which includes a method and apparatus for transmitting specifically requested graphic and/or textual data from an unattended database storage location to a requestor's facsimile machine over a telephone line which includes a host computer such as a PC modified with a facsimile transmission board and a voice generation board. The host computer receives incoming phone calls and prompts the caller using the voice board to select data files by using the DTMF keys of a standard telephone handset. The PC can be left unattended and can run automatically in the facsimile transmission mode. Callers can immediately access needed textual and image data with the use of just a standard telephone and facsimile machine. Multiple workstation nodes can be configured in a network setup to handle a high volume of calls in real time and to allow multiple data services to operate simultaneously.
U.S. Pat. No. 5,165,095 to Mark A. Borcherding describes a method for dialing a telephone, using voice recognition to initiate the dialing and to determine the correct telephone number. The dialing is initiated with a spoken dial command that is recognized by using speaker independent templates that are stored locally with respect to the caller's telephone. The correct telephone number is recognized by using speaker dependent template that are downloaded from a central database or by using speaker independent templates stored locally.
U.S. Pat. No. 5,168,548 to Steven Kaufman et al. describes a reporting system which is disclosed herein, a speech recognizer which is used to select selections of text from a report form stored in a computer and to insert recognized terms in the text hereby to generate a report text under voice control. A command interpreter, also responsive to spoken words, initiates creation of the report text and its subsequent storing, printing and transmission. The command processor is responsive to respective spoken commands to select a destination telephone number and to cause the report text to be sent to apparatus for converting report text to image data and for modulating an audio band signal with the image data for facsimile transmission over telephone lines.
U.S. Pat. No. 5,222,121 to Keiko Shimada describes a voice recognition dialing unit of a telephone mounted on a vehicle or similar mobile body and which allows a call to be originated with ease. When the user of the telephone enters a voice command on voice inputting section, the dialing unit originates a call automatically and thereby connects the other party to the telephone line. In a call origination procedure, the operations for call origination and the verifications are performed between the user and the unit in an interactive sequence. In a preferred embodiment, the unit has a particular call origination procedure in which, when the other party recognized by the unit is wrong as determined by the user by verification, lower place candidates for the other party are called up in response to a particular voice command. In an alternative embodiment, the unit indicates the other party by voicing a name for verification purpose. The alternative embodiment selects and stores only the name of the other party in response to an entered voice signal and, in the event of response for verification, combines the name having been stored and response information stored beforehand to produce composite response voice.
U.S. Pat. No. 5,231,670 to Richard S. Goldbor et al. describes a system and method for generating text from a voice input that divides the processing of each speech event into a dictation event and a text event. Each dictation event handles the processing of data relating to the input into the system, and each text event deals with the generation of text from the inputted voice signals. In order to easily distinguish the dictation events from each other and text events from each other the system and method creates a data structure for storing certain information relating to each individual event. Such data structures enable the system and method to process both simple spoken words as well as spoken commands and to provide the necessary text generation in response to the spoken words or to execute an appropriate function in response to a command. Speech recognition includes the ability to distinguish between dictation text and commands.
U.S. Pat. No. 5,239,586 to Kuniyoshi Marui describes a voice recognition system which comprises a handset and a hands-free microphone for generating an input audio signal, a high-pass filter for eliminating low frequency components from the signal from the handset or hands-free microphone, a signal lever controller for adjusting the level of the high-pass signal in response to the user of either the handset or hands-free microphone, a storer for storing the speech data and a controller for controlling the store so that a user's utterance is stored or the user's utterance is recognized by comparing the utterance to speech data already stored. The handset hook switch provides an on-hook control signal to reduce amplifier gain during hands-free microphone operation.
U.S. Pat. No. 5,301,227 to Shoichi Kamei et al. describes an automatic dial telephone that is useable in a motor vehicle, when a voice input is provided during a period in which input of the names of called parties is awaited, a voice pattern of the name of the called party is compared with reference patterns of called parties stored in reference patterns storing device, to determine the degree of the similarity therebetween. The names of the called parties are output to a user in the order of decreasing degree of similarity. Each time the name of a called party is output, a command word for confirmation is a waited from a user for a predetermined time period. When a voice confirmation command is input and is recognized during this waiting period, a telephone number corresponding to the name of the called party is supplied to a channel. Consequently, the command word for confirmation may be input only if the name of the called party outputted is one desired by the user. Sensors continually monitor the driving condition of the motor vehicle in which the telephone is installed. When the operation of the steering wheel or brakes of the motor vehicle exceeds a predetermined threshold or the speed of the motor vehicle is excessive, the sensors generate safety signals that inhibit the operation of the telephone.
U.S. Pat. No. 5,333,276 to E. Earle Thompson et al. describes a communication system which is provided with multiple purpose personal communication devices. Each communication device includes a touch-sensitive visual display to communicate text and graphic information to and from the user and for operating the communication device. Voice activation and voice control capabilities are included within communication devices to perform the same functions as the touch-sensitive visual display. The communication device includes a built-in modem, audio input and output, telephone jacks and wireless communication. A plurality of application modules are used with personal communication devices to perform a wide variety of communication function such as information retrievable, on-line data base services, electronic and voice mail. Communication devices and application modules cooperate to allow integrating multiple functions such as real time communication, information storage and processing, specialized information services, and remote control of other equipment into an intuitively user friendly apparatus. The system includes both desktop and hand-held communication devices with the same full range of communication capabilities provided in each type of communication device.
U.S. Pat. No. 5,349,636 to Roberto Irribarren describes a communication system for verbal telephonic communication which has a voice message system for storing and retrieving voice messages integrated with a computer database accessing system for storing and retrieving text messages from a separate computer system and for converting the text messages into voice. The systems are integrated via a network which coordinates the functions of each individual system. Additionally, the input/output ports of the voice message system and the computer database accessing system are connected to a parallel fashion to at least one telephone line. In this configuration a user may access both voice messages and database information, including text or electronic mail messages, with a single telephone call. Optionally, facsimile messages can be stored, retrieved and manipulated with a single telephone call.
U.S. Pat. No. 5,406,618 to Stephen B. Knuth et al. describes a telephone answering device that is activated by a proximity sensor when a user crosses its field of detection and whose operation is controlled by simple voice commands. The device incorporates speaker-independent voice recognition circuitry to respond to spoken commands of the user that are elicited by a system generated voice request menu. The telephone answering device performs all the basic functions of a telephone answering machine in response to these simple commands and there is no need for the user to manually operate the telephone answering device.
U.S. Pat. No. 5,602,963 to W. Michael Bissonnette et al. describes a small, portable, hand-held electronic personal organizer which performs voice recognition on words spoken by user to input data into the organizer and records voice messages from the user. The spoken words and the voice messages are input via a microphone. The voice messages are compressed before being converted into digital signals for storage. The stored digital voice messages are reconverted into analog signals and then expanded for reproduction using a speaker. The organizer is capable of a number of different functions, including voice training, memo record, reminder, manual reminder, timer setting, message review, waiting message, calendar, phone group select, number retrieval, add phone number, security and “no” logic. During such various functions, data is principally entered by voice and occasionally through use of a limited keypad, and voice recordings are made and played back as appropriate. A visual display provides feedback to the user. during the various function, the user can edit various different data within the organizer by eliminating or correcting such data or entering new data.
U.S. Pat. No. 5,621,658 to Brion K. Jackson describes an action contained within an electronic mail object which is communicated from a data processing system to another data processing system via an audio device. The action is executable on a data processing system. At the sending data processing system, the action is converted to a predetermined audio pattern. The electronic mail object may contain text in addition to an action. The text is also converted to an audio pattern. The audio pattern are then communicated to the audio device over telephone lines or other communication medium. At the receiving end, the audio device records the object. A user can provide the recorded object to a data processing system, when then executes the action and converts the text audio patterns back to text. In addition, the action can be converted to text and displayed on the data processing system.
U.S. Pat. No. 5,631,745 to John J. Wong et al. describes a telephone terminal adapted for business or home use that includes the ability to receive and send facsimiles, a voice answering function and a computer modem. Various input and output devices may be used for the facsimile function. A voice annotated facsimile may be sent and received. At the same time the facsimile is viewed on a video monitor or ordinary television set, an accompanying voice message is heard through the second system of the monitor or television set. The terminal has an architecture including a central processor and an internal bus structure to which several types of memory, various input-output devices and an interface with the telephone line are connected, among others. Audio Random Access Memory (ARAM) is used for storing both facsimile data and voice data.
U.S. Pat. No. 5,671,328 to Gregory P. Fitzpatrick et al. describes a method and data processing system which are disclosed for automatically creating voice processing template entities. In one embodiment, the invention automatically assembles a plurality of commands received by the data processing system, at least one of said commands having a voice recognition criteria component associated therewith, counts the occurrences of the plurality of commands, assembles voice recognition criteria components associated with the plurality of commands, and, as a result of the occurrence count exceeding a predefined minimum, constructs a voice recognition template entry by associating the assembled voice recognition criteria components with the assembled plurality of commands.
U.S. Pat. No. 5,850,627 to Joel M. Gould et al. describes a word recognition system which can: respond to the input of a character string from a user by limiting the words it will recognize to words having a related, but not necessarily the same, string: score signals generated after a user has been prompted to generate a given word against words other than the prompted word to determine if the signal should be used to train the prompted words; vary the number of signals a user is prompted to generate to train a given word as a function of how well the training signals score against each other or prior models for the prompted word; create a new acoustic model of a phrase by concatenating prior acoustic models of the words in the phrase; obtain information from another program running on the same computer, such as its commands or the context of text being entered into it, and use that information to vary which words it can recognize; determine which program unit, such as an application program or dialog box, currently has input focus on its computer and create a vocabulary state associated with that program unit into which vocabulary words which will be made active when that program group has the focus can be put; detect the available computational resources and alter the instructions it executes in response; test if its ability to respond to voice input has been shut off without user confirmation, and, if so, turn that ability back on and prompt the user to confirm if that ability is to be turned off; store both a first and a second set of models for individual vocabulary words and enable a user to selectively cause the recognizer to disregard the second set of models for a selected word; and/or score a signal representing a given word against models for that word from different word model sets to select which model should be used for future recognition.
Nothwithstanding the prior art, the present invention is neither taught nor rendered obvious thereby.
A voice activated/voice responsive item locator system is disclosed to enable a user to speak into the system and have the system respond with location information for an item requested by the user. For example, shopper at a home supply store may pick up a locator phone or just speak into a wall mounted or otherwise situated microphone and say “Locate Outdoor Paint” or “Find Hammers” of simply state what is sought without the use of a verb, e.g. “Caulking”. The system may reply either with voice or visual (words on a screen, or map), or both voice and visual, e.g. “Aisle 3, Shelf 4”. In some instances the system will reply, for example, with a “Repeat”, or “Restate in different words” or “Please talk to information desk” or other default instructions.
The locator system may be a stand alone device, but in most embodiments would be part of an internal connected system. It could be an intranet or secured internet system, but would in many cases be a storewide system with a plurality of user locations (units, phones, or microphones, with feedback at each location). The system will include an embedded voice-driven interface for speech control of: (1) operational instructions; (2) core system locator function operations, that is, recognition of specific requests and responses thereto; and, (3) optional and default functions. In preferred embodiments, the present invention device is both operated by speech (speech or voice activated) and speech responsive (voice answers and instructions to the user from the system). Thus, the present invention device relies upon automatic speech recognition (ASR), either in place of or in addition to manual locator systems, e.g. books, list, map and computer directories. In some embodiments, user feedback features are included wherein both audio and visual feedback is given to a user in response to recognizable voice signals, while in other possible embodiments, the user may designate audio or visual.
The present invention should be more fully understood when the specification herein is taken in conjunction with the drawings appended hereto wherein:
The present invention is a voice activated/voice responsive item locator and system. By “item” is meant a place or thing that a user desires to locate. Thus, a item could be a particular brand of canned string beans, a type of outdoor stain, a booth at a convention, a particular part in inventory for sale, assemblage or distribution, a particular automobile in a production facility lot or in a large parking garage, or a room, a functional group or a person in an office building or the like. The response may be in the form of a word or sentence presented visually or audibly and it may designate an aisle, a shelf, a bin number, a room number, a row and slot or space, etc.
The voice recognition system digitizes words spoken via a receiver (microphone) handset, headset, or built-in microphone for conversion from analog to digital utilizing a continuous speech recognition digital signal processor (DSP). The main support structure may be a conventional type housing for phones and other communication devices, may be of a different shape or configuration or may be built into a device such as a wall or desk unit, with or without monitor. They could be portable or permanently affixed and could be powered by any means available, e.g. AC or DC current. In the portable mode, the system would be wireless for the user and would, in that respect operate like a cell phone, two way radio, “walkie talkie” or other short distance wireless device, but would have a processor at a central or fixed location having the same features as described above, i.e., the DSP with programming capabilities, etc.
The DSP is connected to a programmable microprocessor and either by customized input or a standard program, the system enables the user to quickly enter voice-activated fields, e.g,, such as “Where is . . . ”, “Find . . . ”, etc. Verification of voice recognition accuracy (prior to execution) is optional and may be accomplished via synthesized voice playback and/or a screen confirmation which requires a “YES” or “NO” to execute or open for revision. In some preferred embodiments, a screen, e.g., LCD, enables visual feedback during input phase, with support for detection, insertion, correction, etc. Cancellation of the entire command or programming instructions may be possible at any time (prior to execution), via keystroke or voice command.
The essential features of the present invention involve the creation of a voice based guide or locator to offer enhanced convenience and speed to users for location of one or more items.
The system may be preprogrammed with the user being required to follow concise instructions for activation and operation, or may be programmable to alter, add or enhance ease or methods of use, e.g. through a limited access code, for manager inputs 3 of user instructions. In any event, manager inputs 3 shall include functional selections and inputs of items and their locations, with provision for subsequent access for modification. This programming may include direct keyboard, voice, etc., and, as mentioned, may include security capabilities for preventing unauthorized use, e.g., voice identification (user recognition) or user security code system, as well as other options which may be included therein, such as a “help” detailed manager instruction section.
Once the system has been programmed for use, the user operation unit(s) 5 provide functional access, which may be passive, i.e., the user speaks, picks up a phone, presses a button, or otherwise takes some action to activate the system; or it may be active, i.e., a proximity sensor, a periodicity timer, or other internal mechanism may automatically activate the system and could trigger an audio or visual query, such as “May I help you locate a product?”
Once the system has been activated and a user has stated the necessary words of input to activate the device, recognition/non-recognition response 7 results from processing the user inputs to central processor 1, and audio and/or video response unit(s) 9 provide feedback 11 to the user, either by answering the inquiry, conditionally defaulting, e.g., asking for a repeat or a restate the question, or fully defaulting e.g. directing the user to a courtesy desk or check out counter for help.
Each acoustic model represents a short sound. The interpreter combines these sounds into words using a dictionary. This dictionary specifies the pronunciation of each word in terms of the acoustic models. After identifying the most likely word, the interpreter then joins sets of models together (using a Viterbi decoder) in a series of pre-defined connections such that paths can be established to provide for a degree of “natural language” recognition; in other words, the user can say “Find hammers”, “Where are hammers” or “hammers” and they are all understood to mean the same thing. Moreover, these sets of models and dictionaries are interchangeable, allowing the same voice recognition component to be used in a variety of applications.
As the voice recognition component is running continuously, there needs to be a way onto distinguish background conversations that might accidentally trigger an unwanted action by the device. For example, two people standing by a voice-activated device might be discussing locations of different goods in a supermarket and be misinterpreted or undesireably resonded to. To avoid this problem, the recognition unit requires a command word to trigger before beginning further recognition. The trigger word is a user-definable setting.
Obviously, numerous modifications and variations of the present invention are possible in light of the above teachings. It is therefore understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein.