WO2004008295A2 - Systeme et procede pour une analyse medicale des caracteristiques de la voix - Google Patents
Systeme et procede pour une analyse medicale des caracteristiques de la voix Download PDFInfo
- Publication number
- WO2004008295A2 WO2004008295A2 PCT/US2003/022636 US0322636W WO2004008295A2 WO 2004008295 A2 WO2004008295 A2 WO 2004008295A2 US 0322636 W US0322636 W US 0322636W WO 2004008295 A2 WO2004008295 A2 WO 2004008295A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- voice
- data
- originator
- captured
- template
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 118
- 238000004458 analytical method Methods 0.000 title claims description 51
- 238000005516 engineering process Methods 0.000 claims description 91
- 238000012512 characterization method Methods 0.000 claims description 23
- 238000003860 storage Methods 0.000 claims description 19
- 230000002068 genetic effect Effects 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 12
- 238000004891 communication Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000012546 transfer Methods 0.000 claims description 8
- 238000004519 manufacturing process Methods 0.000 claims description 7
- 230000010076 replication Effects 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 6
- 230000015654 memory Effects 0.000 claims description 6
- 230000003362 replicative effect Effects 0.000 claims description 6
- 238000013475 authorization Methods 0.000 claims description 3
- 210000000867 larynx Anatomy 0.000 claims description 3
- 230000035479 physiological effects, processes and functions Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 claims description 2
- 238000004590 computer program Methods 0.000 claims 12
- 239000012634 fragment Substances 0.000 claims 8
- 108091028043 Nucleic acid sequence Proteins 0.000 claims 2
- 230000000977 initiatory effect Effects 0.000 claims 2
- 230000003213 activating effect Effects 0.000 claims 1
- 238000013500 data storage Methods 0.000 claims 1
- 230000033764 rhythmic process Effects 0.000 claims 1
- 238000012800 visualization Methods 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 13
- 108020004414 DNA Proteins 0.000 description 24
- 238000011161 development Methods 0.000 description 13
- 230000018109 developmental process Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 9
- 238000010200 validation analysis Methods 0.000 description 9
- 230000015572 biosynthetic process Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 230000007246 mechanism Effects 0.000 description 7
- 230000008901 benefit Effects 0.000 description 6
- 238000009826 distribution Methods 0.000 description 6
- 230000002452 interceptive effect Effects 0.000 description 6
- 108090000623 proteins and genes Proteins 0.000 description 6
- 238000011160 research Methods 0.000 description 6
- 238000003786 synthesis reaction Methods 0.000 description 6
- 230000001755 vocal effect Effects 0.000 description 6
- 241000124008 Mammalia Species 0.000 description 5
- 230000003044 adaptive effect Effects 0.000 description 5
- 230000036541 health Effects 0.000 description 5
- 241000282412 Homo Species 0.000 description 4
- 238000013178 mathematical model Methods 0.000 description 4
- 230000001537 neural effect Effects 0.000 description 4
- 230000037361 pathway Effects 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 210000004556 brain Anatomy 0.000 description 3
- 230000001413 cellular effect Effects 0.000 description 3
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 230000002194 synthesizing effect Effects 0.000 description 3
- 230000002123 temporal effect Effects 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 206010013952 Dysphonia Diseases 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 230000002996 emotional effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 210000004072 lung Anatomy 0.000 description 2
- 238000007620 mathematical function Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 239000002243 precursor Substances 0.000 description 2
- 230000029058 respiratory gaseous exchange Effects 0.000 description 2
- 241000894007 species Species 0.000 description 2
- 230000009885 systemic effect Effects 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 210000001260 vocal cord Anatomy 0.000 description 2
- 230000002747 voluntary effect Effects 0.000 description 2
- 206010013887 Dysarthria Diseases 0.000 description 1
- 206010016275 Fear Diseases 0.000 description 1
- 102100028115 Forkhead box protein P2 Human genes 0.000 description 1
- 101150075185 Foxp2 gene Proteins 0.000 description 1
- 101001059881 Homo sapiens Forkhead box protein P2 Proteins 0.000 description 1
- 229920002153 Hydroxypropyl cellulose Polymers 0.000 description 1
- 238000007476 Maximum Likelihood Methods 0.000 description 1
- 235000002595 Solanum tuberosum Nutrition 0.000 description 1
- 244000061456 Solanum tuberosum Species 0.000 description 1
- 206010044565 Tremor Diseases 0.000 description 1
- 206010000269 abscess Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000011956 best available technology Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013481 data capture Methods 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000003748 differential diagnosis Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 208000001606 epiglottitis Diseases 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 235000010977 hydroxypropyl cellulose Nutrition 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004573 interface analysis Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002483 medication Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 239000003607 modifier Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000000926 neurological effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 229920001690 polydopamine Polymers 0.000 description 1
- 229920000379 polypropylene carbonate Polymers 0.000 description 1
- 238000002300 pressure perturbation calorimetry Methods 0.000 description 1
- 230000035755 proliferation Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000012358 sourcing Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
- G10L17/26—Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
Definitions
- a patient's voice characteristic is frequently an indicator of an unwanted medical condition to a primary care physician, regardless of any prior knowledge of the patient.
- a shortage of physicians and a combined inability of patients to be able to get to physicians and medical care promotes greater interest in telemedicine techniques. Improvements to telemedicine are long overdue, and provide opportunities not previously recognized to general populations.
- a tape or digital recording device is used to record someone's voice and thereby retain it for future listening and replay as it was recorded originally, or portions of the original recording may be played as desired.
- These devices and methods of voice recording also include a range of artificial voices, created by computers, which may be used for many different functions, including for example telephone automatic assistance and verification, very basic speech between toys or equipment and users, synthesized voices for the film and entertainment industry, and the like.
- these artificial voices are preprogrammed to a narrow set of responses according to a specific input.
- these artificial voice sounds are nevertheless simple compared to the robust voice capabilities of the present invention. Indeed, in certain embodiments of the invention there are elements that are either quite different from such systems or which take the previous technology far beyond that ever contemplated or even suggested by such prior discoveries or innovations.
- Figure 1 is a flow diagram of one embodiment of the system operation of the invention.
- Figure 2 is a schematic diagram of one embodiment of a voice capture subsystem.
- Figure 3 is a schematic diagram of one embodiment of a voice analysis subsystem.
- Figure 4 is a schematic diagram of one embodiment of a voice characterization subsystem.
- Figure 5 is a schematic diagram of one embodiment of a voice template subsystem.
- Figure 6 is a schematic diagram of one embodiment of a voice template signal bundler subsystem.
- Figure 7 is one embodiment of a schematic diagram of the system of the invention used with remote information download and upload options.
- Figure 8 is one embodiment of an exemplary plan view of an embodiment of the invention embodied in a mobile, compact component.
- Figure 9 is an exemplary plan view of an embodiment of the invention used with a visual media source.
- Systems and methods are provided for recording or otherwise capturing an enabling amount of a specific person's voice to form a voice pattern template. That template is then useful as a tool for building new speech sounding like that precise voice, using the template, with the new speech probably never having been actually said or never having been said in the precise context or sentences by the specific human but actually sounding identical in all aspects to that specific human's actual speech.
- the enabling portion is designed to capture the elements of the actual voice necessary to reconstruct the actual voice, however a confidence rating is available to predict the limits of the reconstructed or re-created speech in the event there is not enough enabling speech to start with.
- a new voice or voices may be used with a database of subject matter, historical data, and adaptive or artificial intelligence modules to enable new discussions with the user just as if the templated voice's originator were present.
- This system and method may be combined with other media, such as a software file, a chip embedded tool, or other forms. Interactive use of this system and method may occur in various manners.
- a unit module itself may comprise the entirety of an embodiment this invention, e.g. a chip or electronic board which is configured to capture and enable use of a voice in the manner disclosed herein.
- the template is useful, for example, as a tool for capturing and creating new dialogs with people whom are no longer immediately available, who may be deceased, or even those who consent to having the voices templated and used in this manner.
- Another example is the application to media, such as film or photos or other depictions of the actual voice(s) originator to create on-demand virtual dialog with the originator.
- media such as film or photos or other depictions of the actual voice(s) originator to create on-demand virtual dialog with the originator.
- Various other uses and applications are contemplated within the scope of the invention. Detailed Description of the Invention
- Voice is a sound of extraordinary power among mammals.
- the sound of a mother's voice is recognized by and soothes a child even before birth, and the sound of a grandfather's voice calms the fears of even a grown person.
- Other voices may inspire complete strangers or may elicit memories from loved ones of long past events and moments.
- this particularity of one's voice derives from the genetic contribution of the parents resulting in the shape, size, position and development of the various human body components that influence the way one sounds when speaking or otherwise communicating with voice or through the mouth and nasal passages.
- One method of synthesizing voices and sounds is referred to as concatenative, and refers to the recordings of wave form data samples or real human speech. The method then breaks down the pre-recorded original human speech into segments and generates speech utterances by linking these human speech segments to build syllables, words, or phrases. The size of these segments varies.
- Another method of human speech synthesis is known as parametric. In this method, mathematical models are used to recreate a desired speech sound. For each desired sound, a mathematical model or function is used to generate that sound. As such, the parametric method is generally without human sound as an element.
- parametric speech synthesizers there are generally a few well-known types.
- articulatory synthesizer which mathemetically models the physical aspects of the human lungs, larynx, and vocal and nasal tracts.
- formant synthesizer which mathematically models the acoustic aspects of the human vocal tract.
- This invention describes various techniques for templating specific human voices; disclosure regarding development of standardized classification, distribution, and reconstruction modalities for novel human voice templates built with unique voice profile subcodes; and various alternative paths for analyzing and assessing human voice characteristics, and creating very specific patient-related and overall individual person-related voice codes. All of these embodiments are incorporated into the current application.
- the invention comprises developing remote phone-based analyzing modalities of patient voices for telemedicine screening applications.
- the product and service inventions will provide remote identification of certain disease onset via detectable changes in patient voice.
- Patient confidentiality may also be enhanced and provided for using this feature as a biometric validation during a phone call.
- a patient's voice characteristic is frequently an indicator of an unwanted medical condition to a primary care physician, regardless of any prior knowledge of the patient. Examples include detection of tremor, Parkinsonian voice (paucity of expression), psychoactive medications (monotone), peritonsular abscess (known as "hot potato in the throat"), epiglottitis, neurological changes causing dysarthria (changed speech due to brain changes).
- a product or service utilizing this technology will provide instantaneous analysis and recurrent updates of a patient's voice, compare the updates to a critical code which defines a patient baseline voice (when the patient is known), and generate a differential diagnosis based on likely conditions triggering certain anomalies to that patient's voice, whether the patient is a first time caller or not.
- a digital voice code with ultra-high patient specificity, and novel medical condition recognition capabilities (preferably) by digital signal analyses are provided and disclosed herein.
- the digital voice code is derived from any of the plurality of techniques and algorithms, as disclosed herein below.
- the medical condition recognition capabilities and algorithms are those being developed by Applicant in relation to the above-listed conditions and others.
- this technology also provides additional business method opportunities.
- utilization of this technology in a blind voluntary automated screening system for an organization-communication system with private queuing to voice originators would allow for non-traditional intervention and courtesy alerts to non- patient participants in the plan, as well as traditional patients in order to screen for major disease onsets and have wider organizational impact on healthcare savings.
- Use by large non-medical organizations as a daily screening tool with their phone systems is a very likely utilization of this technology, in addition to telemedicine call-in applications utilized within healthcare organizations.
- this technology has not been possible because the enabling voice platform technology and the medical condition recognition capabilities and algorithms were not created.
- This telemedicine approach is particularly unique in its use of the digital code for each person's voice, which allows extensive data comparison and record checking capabilities for each patient as well.
- Research indicates a very large population which is readily reachable and is also affected by the main diseases/conditions currently being investigated for use by this technology.
- Other applications exist as well, for example, status post-laryngectomy, or no vocal cords, may use this technology as an alternate to the currently used esophageal speech or vibrator, and then be classified accordingly.
- One goal of this technology is to create a uniform classification system which captures, defines, and quantifies all parameters of a particular human voice.
- a specific digital code will be generated to uniquely identify and define a particular human's voice, in all regards.
- This unique voice identifying code will be useful in later generation of a machine spoken voice using text or other data when combined with the code for that selected human's voice.
- an important element of this technology is the development of standardized classification, distribution, and reconstitution modalities for novel human voice templates built with unique voice profile sub-codes, as disclosed herein.
- the classification system will capture and normalize the results of different voice analysis techniques. These techniques are drawn from voice recognition, voice synthesis/production, text-to- speech know how, and other fields.
- Various analytical tools will be categorized from both the traditional analytical realm (e.g., vocal fold analysis, glottal pulse response, histogram/spectrogram analysis, vocal tract dimensions, fractal analysis, audio-visual analysis, short time Fourier transform, etc.) as well as more novel areas (e.g., genetic characteristics, age, accent, education, health, etc.).
- This voice identifying and classifying system will yield a unique, data-rich descriptive code for each human voice, and such code will function as a cipher key to generate a synthetic voice having identical sound and other characteristics as the original human voice.
- Each code will then be formatted in various ways to facilitate it's use as a building block in licensing or other commercial applications, including electronic transfer and storage. Such a system will enable dramatic changes to the human interface field as well as other technical areas.
- One objective of this technical effort is to identify the best methodology for classifying all elements of a specific voice and then arranging a logical and unique classification system for future standard categorization protocols, which require development. It is well known that numerous technical approaches to voice recognition, authentication, generation, biometrics, and digital communications have been developed and successfully implemented. Our technology will use many of these underlying technical approaches as building blocks to develop voice classification, distribution and re-creation of actual voice profiles. Analysis of an enabling portion of an actual speaker according to a very extensive parsing methodology produces a digital code which uniquely defines the various elements of the speaker's voice. This digital code is referred to as a "voice DNA" and becomes the basis to the particular speaker's voice descriptive code. The diverse yet excellent work done in the voice recognition field is the source of best available technology in this regard. However, that technology has been developed with considerable constraints, and has not focused on the underlying voice DNA of speakers. Rather, it has been more content-dependent on the whole.
- a third technical objective and embodiment is to identify the uses of specific genetic codes and/or gene sequence codes to assist in the unique classification of a particular speaker's voice and to enhance the fidelity of any voice re-created using a voice DNA product. This has always been considered (by us) to be an important tool for classifying the enabling portion of the original voice. We believe this feature alone merits research support, particularly in light of recent discoveries involving the FOXP2 gene, but in this context is considered to be one part of a larger effort to define the best techniques to achieve unique classification of each human's voice. An additional opportunity exists because a derivative of understanding and classifying the mechanisms of individual speech is improved understanding of the specific brain circuitry that underlies such faculty in each person.
- the cognitive information processing role per subject, is a basis of the interpretation and formation of speech. This technical effort will endeavor to utilize this basic element as a further enhancement of the specificity of each person's voice DNA, mindful of identifying future opportunities for research in communication among human-human and human-machine neural interfaces.
- a fourth technical objective is the distribution, validation and control over use of voice DNA. Identifying the best format for these voice classification signals, and the various interface requirements of the reconstituting data, are essential technical milestones. The reality of fusing numerous underlying voice technologies requires technical reconciliation and normalization, at certain levels. The resulting integration will create a new technology field, with considerable revenue and scientific spin-offs. It is believed that a driver of revenue will be the ability to distribute voice DNA or templates via digital networks and the like for use with different products and services. This will require proper validation, transfer and proliferation/derivative rights control protocols and regimes.
- Voice template technology will create revenue from these and other products: chip logic; PC cards; HPCs/PPCs/PDAs; cellular phone systems; travel and ID cards and related biometric systems; internet transfers; voice use license fees; and numerous interface applications within many industries.
- the potential for unforeseen spin-off applications for this technology is also quite high.
- Systems and methods are proposed to capture and enhance an enabling portion of an individual's voice and then to create a voice template or profile signal which may be combined at a later time with noise/text of another origin to utilize the original voice. Such voice may then be used to speak any form/content/text provided via digital or other input thereto, and to "speak" content which was not spoken in an original form by the original voice.
- Products and processes for online use are planned, as are certain business methods and industry applications. Specific objectives guide the technology development, as follows:
- One goal is to identify the best methodology for classifying all elements of a specific voice into a logical and unique code utilizing standard categorization protocols, which require development.
- the particularity of voice derives from the genetic contribution of the parents resulting in the shape, size, position and development of the various human body morphology that influences the way one sounds when speaking or otherwise communicating with voice or through the mouth and nasal passages. Other influences exist as well. It is understandable therefore that there is a range of differences among people, often even within the same family. Indeed, even the same person may sound slightly different according to temporal influences such as the health, stress level, emotional state, fatigue, the ambient temperature around the person, or other factors. Regardless, the ability of humans to associate through their senses is remarkable, particularly as such sensing relates to identification and association with the human voice.
- One method of synthesizing voices and sounds is referred to as concatenative, and refers to the recordings of wave form data samples or real human speech. The method then breaks down the pre-recorded original human speech into segments and generates speech utterances by linking these human speech segments to build syllables, words, or phrases. The size of these segments varies.
- Another method of human speech synthesis is known as parametric. In this method, mathematical models are used to recreate a desired speech sound. For each desired sound, a mathematical model or function is used to generate that sound. As such, the parametric method is generally without human sound as an element.
- parametric speech synthesizers there are generally a few well-known types.
- articulatory synthesizer which mathematically models the physical aspects of the human lungs, larynx, and vocal and nasal tracts.
- formant synthesizer which mathematically models the acoustic aspects of the human vocal tract.
- Other systems include means for recognizing a specific voice, once the using system has been trained in that voice. Examples of this include the various speech recognition systems useful in the field of capturing spoken language and then translating those sounds into text, such as with systems for dictation and the like. Other speech related systems concern the field of biometrics, and use of certain spoken words as security codes or ciphers. What has long been overlooked, however, is a system and method for preserving the key or cipher to voices of other beings in a dynamic and adaptive manner for future use and benefit by the originator or by others.
- Table I is one of several uncorrelated tables which identifies technically relevant characteristics ( C ) of human voice, and tools/technology ( T ) having known analytical value.
- An initial task is to validate these Tables and then to rank best T 1_x for each C.
- the stability of each ranking will need testing against various C categories. For example, whether T 1 is a reliable and stable choice for all voice types or all voices under stress.
- a correlation ranking among selected T categories per C category will also be done. This will result in the best techniques for deriving each characteristic, and then lead to a classification system and reconciliation algorithms suitable for fusing the output of the best techniques into an integrated classification code.
- a master topographic selection model is currently preferred due to its intuitive adaptability to voice dynamics, somewhat similar to that of Genetic Algorithms. Use of GAs and other techniques will likely aid in more rapid assessment and selection of best models.
- Figure 1 is a schematic diagram of one embodiment of a system 10 for capturing an enabling portion of a specific voice sufficient for using that portion as a template in further use of the voice characteristics.
- System 10 may be part of a handheld device, such as an electronic handheld device, or it may be part of a computing device of the size of a laptop, a notebook, or a desktop, or system 1 may be part of merely a circuit board within another device, or an electronics component or element designed for temporary or permanent placement in or use with another electronic element, circuit, or system, or system 10 may, in whole or in part, comprise computer readable code or merely a logic or functional circuit in a neural system, or system 10 may be formed as some other device or product such as a distributed network-style system.
- system 10 comprises input or capture means 15 for capturing or receiving a portion of a voice for processing and construction of a voice algorithm or template means 19, which may be formed as a stream of data, a data package, a telecommunications signal, software code means for defining and re-generating a specific voice, or a plurality of voice characteristics organized for application to or template on another organization of sound or noise suitable to arrange the sound or noise as an apparent voice of an originator's voice.
- Other means of formatting computer readable program code means, or other means, for causing use of certain identified voice characteristics data to artificially generate a voice is also contemplated within this invention.
- the logic or rules of the algorithm or template means 19 are preferably formed with a minimum of voice input, however various amounts of voice and other data may be desired to form an acceptable data set for a particular voice.
- an enabling portion of a human voice for example, with a small amount of analog or digital recording, or real-time live input, of the person's voice that is to be templated.
- a prescribed grouping of words may be formed to optimize data capture of the most relevant voice characteristics of the person to enable accurate replication of the voice.
- Analysis means are contemplated for most efficiently determining what form of enabling portion is best for a particular person. Whether by a single data input or a series of inputs, the voice data is captured and stored in at least one portion of storage means 22.
- a template of the voice is, in one embodiment, stored until called for by the processor means 25. For example, after voice AA has had an enabling portion captured, analyzed and templated (now referred to as AA) it is stored in a storage means 22 (which may be either resident near the other components or located in a remote or distributed mode at one or more locations) until a demand request occurs.
- a demand request is a user of system 10 submitting a request via representative input means 29 to utilize the voice AA template AA t in a newly created conversation with voice AA participating as a generated voice rather than an actual, live use of voice AA.
- This may occur in conjunction with or utilization of one or more various databases, a few of which are represented by situational database 33 or personal database 36.
- voice AA template AA t is called and provided as a forming mechanism with certain other noise to create a new conversational voice AA 1 that sounds precisely like the original voice AA of the originally inputted data, once formed.
- the new voice AA 1 sounds like original voice AA in all respects, it is actually an artificially created voice with the template AA t providing the matching key, such as a genetic code, to voice AA.
- an enabling portion of an actual voice may encode the system 10 using a template to allow regeneration and unlimited utilization of the captured voice in virtually any way desired by the user.
- This is not simply a synthesis of prior utterances of bits of voice AA which are electronically fused together, by either concatenation or formant techniques, but rather an entirely new voice that is designed, manufactured and assembled or constructed using the voice data characteristics of voice AA (i.e., the voice template or profile), and possibly other characteristics relevant to the originator of voice AA, e.g. genetic code, tissue DNA applicable to a specific voice, or other physiologic precursor.
- connection means 41 represents pathways for energy or data flow which may be actual leads, light channels, or other electronic, biologic or other activatable paths among system components.
- power means 44 is shown within system 10, but may also be remote if desired.
- the algorithm, signal, code means or template which is created in whole or in part may be returned for storage or refinement within either storage means 22, template means 19, or other system component or architecture.
- This capability permits and facilitates improvement or adaptation of the specific voice template according to the instructions of the creator or another user. This could be accomplished, for example, if multiple data sets of the same person's voice could be inputted over time, or if different ages, development, or other changes to physiology or temperament of the originator of the voice occur. Indeed, it is possible to train the templated voice to recall the context of previous engagements and to include such knowledge in future operations.
- voice AA 1 template AA ⁇
- voice or template AA ⁇
- location of a person with a voice BB that comprises one or more voice characteristics that are similar to voice AA which was the originator for voice template AA ⁇ .
- voice BB it may be useful to input the one or more similar characteristics from voice BB as either limited or general refinement inputs to voice AA 1 or voice template AA
- voice BB it is then possible to also retain voice BB and create a voice BB 1 and voice template BB ⁇ , either of which may be useful at a future date.
- Another example includes creation of a database of variously refined voices for a single originator of the voice, useful on demand or as appropriate by system or user, according to the situation that is presented.
- a service may be offered to voice match and provide suitable refinement tools, such as natural or artificially generated waveforms or other acoustic or signal elements, to refine voice templates according to the user's desires.
- any use of a voice-like noise which is generated by data provided to and data resulting from a template or coding tool for creation of that voice-like noise, is captured within the scope of this invention, particularly when such coding tool is used with other noise or sound generating means, if needed, to re-create a voice sound that is virtually identical to the originator's actual voice.
- the ability to provide machine, component, or computer readable code means as part of the signal forming or transmitting of the voice template process or product further facilitates use of this technology.
- Means to tie or activate use of this voice templating and voice generating technology to streaming or other forms of data allows for virtual dialog, which may be adaptive and intelligent, as well as merely informational or reactive, and with such dialog or conversations being with voices selected by the user. It is also recognized that the technology herein disclosed may be utilized with visual images as well as aural sounds.
- a voice template as described herein may be created using data that does not include an actual enabling portion of an originator's voice, but that the enabling portion of the originator's voice may be used, possibly with other data, to validate the replication accuracy of the originator's voice.
- a templated or replicated voice may be used to interact with or prompt users of computers or other machines and systems. The user may select such templated voice from either her own library of templated voices, another source of templated voices, or she may simply create a new voice.
- templated voice AA 1 may be selected by the user for voicemail prompts or reading of texts, or other communication interface, whereas templated voice CC may be selected for use in relation to an interactive entertainment use.
- Troubleshooting or problems lurking in the user's machine, or alerting signals to a user of a device, may be identified or resolved by the user while working with templated voice DD.
- Template selection and use, and generated voice creation and use may be accomplished either within the user's machine or device, partially within the user's machine or device, or external of the user's machine or device.
- a parent desired her child to learn about race relations in the United States in the decade of the 1960s using one of the child's deceased grandparent's voices
- the templated voice of the selected grandparent would be designed, manufactured and designated for use.
- System 10 would access one or more databases to harvest information and knowledge about the designated topic and provide that information to one or more databases within system 10, such as situational database 33 for use as needed.
- the grandparents' templated voice EE 1 would be used, following access to the desired information, and the demand request would be met by the templated voice EE 1 commencing a discussion on the designated topic when desired.
- Such discussion can be saved for later use within system 10 or at a remote location as desired, or the discussion may be interactive between the "grandparent" i.e. the templated voice, and the child.
- This feature is possible by use of a voice recognition module to know in advance of the discussion the identity of the child's voice and to include adequate vocabulary and neural cognition of the various question combinations likely from the child.
- a bridge would be provided from the input and voice recognition module to the templated voice portion of the system, to enable responsiveness by the templated voice.
- Various speech recognition tools are conceivable for use in this manner, when so configured according to the novel uses described herein. Of course this configuration also requires means to rapidly search for the answer to the question and to formulate a response appropriate to the listening child.
- this example illustrates the extraordinary potential of this technology, particularly when combined with suitable data, system power, and system speed.
- the optional voice recognition module it is possible to utilize only limited features to enable a listener of a templated voice to direct the generated voice to cease or continue, or to enable certain other features with certain commands. This would be a form of limited interactive mode appropriate for some but not all types of use. Even if the user chose not to use the optional features and instead merely arranged for a story or a discussion in the absent grandparents' voice, the effect and utility of this is enormous to this or other types of uses.
- the templated voice may again be that of the grandparent selected above (templated voice EE 1 ), and the filter of DATA DATES is used with a selected date of "BEFORE DECEMBER 1963" for a discussion of race relations in the United States in the decade of the 1960s. The result would be a discussion that would not include any information that occurred after the designated date. In this example, the "grandparent" could not discuss the Voting Rights Act of 1965 or the urban riots of the late 1960s in that country.
- a user may direct a templated voice of a loved one or someone else to read to the user.
- people of all ages it is possible for people of all ages to have books read to them in the voice of an absent or deceased family member or other person known to the user.
- this innovation alone will provide enormous benefit to users.
- This type of use has wide applications beyond the specific example just provided. Indeed, an even broader use of this technology in this manner is to have available a database of authorized and templated voices which may be accessible and useable by others for a fee or other form of compensation.
- this technology When used for music, this technology has similar profound implications, particularly if one can access templated voices of past and present singers of renown- many of whose voices are still available for templating. Clearly, this technology enables a new industry of manufacturing, leasing, purchasing, or otherwise using voice templates and associated means, techniques and methods of conducting business therewith.
- the invention may also have utility in medical treatments for certain minor or major psychological ailments, for which proper use of templated voice therapy may be quite palliative or even therapeutic.
- Yet another possible use of this technology is to create a newly designed voice for use, but one which has a basis or precursor in one or more templated voices from actual mammalian origin. Ownership and further use of the newly created voice may be controllable under various means or legal enforcement, such as licensing or royalties and the like. Of course, such voices may be retained as private possessions for limited use by the creator as well.
- Such voices will represent the creative aspirations of the creator, but each voice will actually have a component or strain of actual mammalian voice as a basis through use of the templating tool or code, similar to a strand of tissue DNA but applicable to a specific voice.
- This type of combination presents powerful new communication capabilities and relationships based on voice and other sounds created by mammals.
- Systems according to the invention may be handheld or of other size. Systems may be embedded in other systems or may be stand alone in operation. The systems and methods herein may have part or all of the elements in a distributed, network or other remote system of relationship. Systems and methods herein may utilize downloadable or remotely accessible data, and may be used for control of various other systems or methods or processes. Embodiments of the invention include exposed interface routines for requesting and implementing the methods and operations disclosed herein but which may be carried out in whole or in part by other operating or application systems. The templating process and the use of templated voices may be accomplished and used by either mammals or artificial machines or processes. For example, a bot or other intelligent aide may create or use one or more templated voices of this type.
- Such an aide may also be utilized to search for voices automatically according to certain general or limited criteria, and may then generate templated voices in voice factories, either virtual or physical. In this manner, large databases of templated voices may be efficiently created. In this or similar systemic use, it may be desirable to create and apply data or other types of tagging and identification technology to one or more portions of the actual voice utilized to create a templated voice.
- a templating process using elements of the embodiments herein yields a voice coding signal, comprising the logic structure of characteristics of a specific voice essential for accurately replicating the sound of that voice.
- Example 4 A home energy monitor, reporter, or mate, using one or more selected voices using the technology herein.
- Example 4 A home energy monitor, reporter, or mate, using one or more selected voices using the technology herein.
- a hotel room assistant, or automobile assistant to prompt the user according to desired prompting such as for example a wake-up call in a hotel in the voice selected by the user.
- desired prompting such as for example a wake-up call in a hotel in the voice selected by the user.
- an operator of a vehicle might receive information in the voice or voices selected by the user.
- voice template technology in combination with other visual media, such as with a photograph, digital video or a holographic image.
- a personal device that scans and updates downloadable information for a user as desired in voice or voices of one's choosing. For example, this may be useful for organizing actions capable of being done by a bot, such as an info-bot for background searching and interface while the user is not available and then reporting status to the user in one or more designated voices using the technology herein.
- a bot such as an info-bot for background searching and interface while the user is not available and then reporting status to the user in one or more designated voices using the technology herein.
- a safety reminder when used with one or more components of gear or equipment in the workplace, such as a personal computer posture monitor, electrical equipment, dangerous equipment, etc.
- voice activated systems such as dictation devices, as prompts, companions, or text readers.
- Using the technology disclosed herein use as social mediation or control mechanisms, such as a tool against road rage or other forms of anger and frustration, activatable by driver or automatically, or by other means.
- VoiceSelectTM brand of movie or video match technology uses as a VoiceSelectTM brand of movie or video match technology to utilize preferred voices for templating of entertainment script already used by the original performer or subsequently created for voice template technology combination uses.
- an "alter ego” device such as a handheld unit which engages on "SelectVoiceTM” brand or “VoiceXTM” brand mode(s) of operation and has a database of images of those who match the voice as well as anonymous models which can be selected, similar to that referred to in Example 7.
- Using the technology disclosed herein use as a bedtime reader or a night mate in a dwelling for monitoring and interactive security.
- Figure 2 is a flow diagram of one embodiment of a voice capture subsystem which may comprise computer readable code means or method for accomplishing the capture, analysis and use of a voice AA designated for templating.
- Figure 3 is one embodiment of a voice analysis subsystem which may comprise logic or method means for efficiently determining voice data characterization routing.
- voice AA is captured in acquisition module or step 103 and then routed by logic steps and data conductive pathways, such as pathway 106, through the templating process. Capture may be accomplished by either digital or analog methods and components.
- the signal which then represents captured voice AA is routed through analysis means 111 or method to determine whether an existing voice profile or template matches voice AA.
- This may be accomplished, for example, by comparing one or a plurality of characteristics (such as those shown in voice characterization subsystem 113 of Figure 4) as determined by either acquisition module 103 or analysis means 111, and then comparing those one or more characteristics with known voice profiles or templates available for access, such as at analysis step 111.
- Representative feedback and initial analysis loop 114 facilitates these steps, as does pathway 116.
- Such comparison may include querying of a voice profile database or other storage medium, either locally or remotely.
- the analysis step at analysis module 111 and voice characterization subsystem 113 may be repeated according to algorithmic, statistical or other techniques to affirm whether the voice being analyzed does or does not relate or match an existing voice profile or data file.
- Figure 4 provides further detail of voice characterization subsystem 113.
- the signal corresponding to voice AA does not have a match or is not identified with an existing voice profile set then the signal is routed to the voice characterization subsystem for comprehensive characterization.
- creation of a template may not be required at module/step 127.
- the signal might be analyzed and/or characterized for possible generation of a revised profile or template- which itself may then be stored or applied. This situation might occur, for example, when additional characterization data is available (such as size of enabling portion, existence or lack of stress, or other factors) which had not been previously available.
- a specific voice data file might comprise a plurality of templates.
- the template creation module/step 127 of Figure 2 comprises utilizing the voice characterization subsystem to create a unique identifier, preferably a digital identifier, for that specific voice being templated or profiled.
- This data is similar, in the abstract, to genetic codes, gene sequence codes, or bar codes, and like identifiers of singularly unique objects, entities or phenomena. Accordingly, applicants refer to this voice profile or template as "Voice Template TechnologyTM” as well as “Voice DNATM or VDNATM” and “Voice Sequence CodesTM or Voice Sequence CodingTM”.
- the terms "Profile, Profiles or Profiling” and derivative terms may be substituted in the above trademark or other reference terms for this new technology.
- the voice template may be stored (shown at storage module or step 161 or applied in use at module or step 164).
- Figure 4 is a schematic representation of a voice characterization subsystem.
- This disclosure comprises at least one embodiment of characterization data and means for determining and characterizing salient data to define a voice using voice templating or profiling, as disclosed herein.
- various types of data is available for comparison in formulating the characterization data.
- This characterization data will then be used to create the voice template or profile according to coding criteria.
- the data in Figure 4 appears to be arranged in discreet modules, an open comparator process may be preferred in which any data may be accessed for comparison in any of various sequences or weighted priorities.
- data may comprise the categories of language, gender, dialect, region, or accent (shown as "Voice Characteristics" output signal VC 0 at module or step 201); frequency, pitch, tone, duration, or amplitude (shown as output signal V at module or step 203); age, health, pronunciation, vocabulary, or physiology- either genetic or otherwise (shown as output signal VC 2 at module or step 205); patterns, syntax, volume, transition, or voice type (shown as output signal VC 3 at module or step 207); education, experience, phase, repetition, or grammar (shown as output signal VC at module or step 209); occupation, nationality, ethnicity, custom or setting (shown as output signal VC 5 at module or step 211); context, variances, rules/models, enabling portion type, size or number (shown as output signal VC 6 at module or step 213); speed, emotion, cluster, similarities, or acoustic model (shown as output signal VC 7 at module or step 215); math model
- VC X encompasses any known categorization technique at the time of interpretation, regardless of mention herein, provided it is useful in then defining a unique voice profile or template for a specific voice- and is used according to the novel teachings disclosed herein.
- data combined in voice characteristic files and output signals VC 0 , VC b VC 2 , VC 3 , VC 4 , VC 5 , VC 6 , VC 7 , VC 8 , VC 9 , VC 10 , VC U , VC 12 , and VC X may be prioritized and combined in various ways in order to accurately and efficiently analyze and characterize a voice, with VC X representing still further techniques incorporated herein by reference.
- Another goal of this technology is to identify protocols for reconstituting encoded voice DNA with content into a voice recognizable as an originator.
- the voice DNA will be combined with text or similar input data in a manner that is likely similar to compiling steps in known text-to-speech synthesizers.
- a key difference is the addition of the very specific voice DNA template, which functions as a unique recipe or filter for each generated voice.
- Figures 5 and 6 illustrate an exemplary signal bundler suitable for receiving the various voice characteristic data, such as digital or coded data representative of the information deemed relevant and formative of the voice being templated.
- the signal bundler 316 then combines the output of signal content module or step 332 and values/scoring from one or more signals VC 0 - VC X and formats the signal or code at module or step 343 as appropriate for proper transfer and use by various potential user interfaces, devices or transmission means to create an output voice template, code, or signal VT X . It is recognized that various methods are possible to create a unique identifier to delineate the various voice characteristics- and that such various possibilities are enabled herein in view of the broader context and scope of this invention- to a certain degree independent of some component methodology.
- Another goal of this technology is to identify the use of specific genetic codes and/or gene sequence codes to assist in the unique classification of a particular speaker's voice, and to enhance the fidelity of any voice re-created using a voice DNA product.
- Another goal is to determine the best methodology for distribution, validation and use-control of voice DNA. It is recognized, of course, that the implications of this technology are vast, and safeguards will be necessary to maintain the proper use of this templated voice technology. Indeed, this technology may require further use of authorization protocols to only allow authorized users to access and use the voice template technology and data. An additional necessity may be to have mechanisms for verifying that voices heard are either real or templated, in order to ensure against fraudulent or unauthorized use of such created voices.
- Systems using this technology effort may be handheld or of other size.
- Systems may be embedded in other systems or may be stand alone in operation.
- the systems and methods herein may have part or all of the elements in a distributed, network or other remote system of relationship.
- Systems and methods herein may utilize downloadable or remotely accessible data, and may be used for control of various other systems or methods or processes.
- Embodiments of the technology include exposed interface routines for requesting and implementing the methods and operations disclosed herein but which may be carried out in whole or in part by other operating or application systems.
- the templating process and the use of templated voices may be accomplished and used by either mammals or artificial machines or processes. For example, a bot or other intelligent aide may create or use one or more templated voices of this type.
- Such an aide may also be utilized to search for voices automatically according to certain general or limited criteria, and may then generate templated voices in voice factories, either virtual or physical. In this manner, large databases of templated voices may be efficiently created. In this or similar systemic use, it may be desirable to create and apply data or other types of tagging and identification technology to one or more portions of the actual voice utilized to create a templated voice or to use voice tags for marking other data.
- Figure 7 is a representative organization and method of an electronic query and transfer between a voice template generation or storage facility 404 and a remote user.
- enabling portions may be sent to a remote voice template generation or storage facility 404 by any number of various users 410, 413, 416.
- the facility 404 then generates or retrieves a voice template data file and creates or retrieves a voice template signal.
- the template signal is then transmitted or downloaded to the user or its designee, shown at step 437.
- the template signal is formatted for appropriate use by a destination device, including activation instructions and protocols, shown at step/module 457.
- Figure 8 is a schematic representation of a mobile medium, such as a card, disk, or chip on which are essential components, depending on the user mode and need, for utilizing voice template technology.
- a hotel door card 477 may be provided at check-in to a hotel by a traveler.
- additional features incorporating aspects of this invention may be made available.
- a schematic representation of optional features within such a card include means 481 for receiving and using a voice template for a voice or voices selected by the traveler for various purposes during the traveler's stay at the hotel.
- such features may include a template receiving and storage element 501, a noise generator or generator circuitry 506, a central processing unit 511, input output circuitry 515, digital to analog/analog to digital elements 518, and clock means 521.
- various other elements may be utilized, such as voice compression or expansion means- such as those known in the cellular phone industry, or other components to enable the card to function as desired.
- voice compression or expansion means such as those known in the cellular phone industry, or other components to enable the card to function as desired.
- the user may then enjoy dialog or interface with inanimate devices within the hotel in the voice(s) selected by the traveler. Indeed, a traveler profile may even retain such voice preference information, as appropriate, and certain added billings or benefits may accrue through use of this invention. It is recognized that the invention may be employed in a wide variety of applications and articles, and the example of Figures 8 and 9 should not be considered limiting.
- Figure 9 is a depiction of a photograph 602 which is configured for interactive use of voice template technology with voice JJ attributable to figure F JT and voice KK attributable to figure F KK - Means are combined with the frame 610 or other structure, whether computer readable code means or simple three dimensional material, for interfacing the subjects or objects of the photo (or other media) with the appropriate voice templates to recreate a dialogue that either likely occurred or could have occurred, as desired by the user.
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Medical Treatment And Welfare Office Work (AREA)
- Machine Translation (AREA)
Abstract
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2003259177A AU2003259177A1 (en) | 2002-07-17 | 2003-07-17 | System and method for voice characteristic medical analysis |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US39670202P | 2002-07-17 | 2002-07-17 | |
US60/396,702 | 2002-07-17 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2004008295A2 true WO2004008295A2 (fr) | 2004-01-22 |
WO2004008295A3 WO2004008295A3 (fr) | 2004-04-15 |
Family
ID=30116053
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2003/022636 WO2004008295A2 (fr) | 2002-07-17 | 2003-07-17 | Systeme et procede pour une analyse medicale des caracteristiques de la voix |
Country Status (2)
Country | Link |
---|---|
AU (1) | AU2003259177A1 (fr) |
WO (1) | WO2004008295A2 (fr) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105448291A (zh) * | 2015-12-02 | 2016-03-30 | 南京邮电大学 | 基于语音的帕金森症检测方法及检测系统 |
CN110211566A (zh) * | 2019-06-08 | 2019-09-06 | 安徽中医药大学 | 一种基于压缩感知的肝豆状核变性言语障碍的分类方法 |
CN110223688A (zh) * | 2019-06-08 | 2019-09-10 | 安徽中医药大学 | 一种基于压缩感知的肝豆状核变性言语障碍的自评估系统 |
CN111326162A (zh) * | 2020-04-15 | 2020-06-23 | 厦门快商通科技股份有限公司 | 一种声纹特征的采集方法和装置以及设备 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5007081A (en) * | 1989-01-05 | 1991-04-09 | Origin Technology, Inc. | Speech activated telephone |
US5594789A (en) * | 1994-10-13 | 1997-01-14 | Bell Atlantic Network Services, Inc. | Transaction implementation in video dial tone network |
US5717828A (en) * | 1995-03-15 | 1998-02-10 | Syracuse Language Systems | Speech recognition apparatus and method for learning |
US5774841A (en) * | 1995-09-20 | 1998-06-30 | The United States Of America As Represented By The Adminstrator Of The National Aeronautics And Space Administration | Real-time reconfigurable adaptive speech recognition command and control apparatus and method |
-
2003
- 2003-07-17 WO PCT/US2003/022636 patent/WO2004008295A2/fr not_active Application Discontinuation
- 2003-07-17 AU AU2003259177A patent/AU2003259177A1/en not_active Abandoned
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5007081A (en) * | 1989-01-05 | 1991-04-09 | Origin Technology, Inc. | Speech activated telephone |
US5594789A (en) * | 1994-10-13 | 1997-01-14 | Bell Atlantic Network Services, Inc. | Transaction implementation in video dial tone network |
US5717828A (en) * | 1995-03-15 | 1998-02-10 | Syracuse Language Systems | Speech recognition apparatus and method for learning |
US5774841A (en) * | 1995-09-20 | 1998-06-30 | The United States Of America As Represented By The Adminstrator Of The National Aeronautics And Space Administration | Real-time reconfigurable adaptive speech recognition command and control apparatus and method |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105448291A (zh) * | 2015-12-02 | 2016-03-30 | 南京邮电大学 | 基于语音的帕金森症检测方法及检测系统 |
CN110211566A (zh) * | 2019-06-08 | 2019-09-06 | 安徽中医药大学 | 一种基于压缩感知的肝豆状核变性言语障碍的分类方法 |
CN110223688A (zh) * | 2019-06-08 | 2019-09-10 | 安徽中医药大学 | 一种基于压缩感知的肝豆状核变性言语障碍的自评估系统 |
CN111326162A (zh) * | 2020-04-15 | 2020-06-23 | 厦门快商通科技股份有限公司 | 一种声纹特征的采集方法和装置以及设备 |
Also Published As
Publication number | Publication date |
---|---|
WO2004008295A3 (fr) | 2004-04-15 |
AU2003259177A1 (en) | 2004-02-02 |
AU2003259177A8 (en) | 2004-02-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20020072900A1 (en) | System and method of templating specific human voices | |
Fagherazzi et al. | Voice for health: the use of vocal biomarkers from research to clinical practice | |
Bachorowski | Vocal expression and perception of emotion | |
Rachman et al. | DAVID: An open-source platform for real-time transformation of infra-segmental emotional cues in running speech | |
Rao et al. | Robust emotion recognition using spectral and prosodic features | |
Kraljic et al. | First impressions and last resorts: How listeners adjust to speaker variability | |
Arias et al. | Beyond correlation: acoustic transformation methods for the experimental study of emotional voice and speech | |
CN108132995A (zh) | 用于处理音频信息的方法和装置 | |
Livingstone et al. | Common cues to emotion in the dynamic facial expressions of speech and song | |
CN113010138B (zh) | 文章的语音播放方法、装置、设备及计算机可读存储介质 | |
US20050108011A1 (en) | System and method of templating specific human voices | |
Caponetti et al. | Biologically inspired emotion recognition from speech | |
AU2048001A (en) | System and method of templating specific human voices | |
Hashem et al. | Speech emotion recognition approaches: A systematic review | |
Lotfian et al. | Lexical dependent emotion detection using synthetic speech reference | |
He | Stress and emotion recognition in natural speech in the work and family environments | |
Wu et al. | Exemplar-based emotive speech synthesis | |
Garcia-Cuesta et al. | EmoMatchSpanishDB: study of speech emotion recognition machine learning models in a new Spanish elicited database | |
Han et al. | [Retracted] The Modular Design of an English Pronunciation Level Evaluation System Based on Machine Learning | |
WO2004008295A2 (fr) | Systeme et procede pour une analyse medicale des caracteristiques de la voix | |
Fürer et al. | Supervised speaker diarization using random forests: a tool for psychotherapy process research | |
Potapova et al. | Forensic identification of foreign-language speakers by the method of structural-melodic analysis of phonograms | |
Ramati | Algorithmic Ventriloquism: The Contested State of Voice in AI Speech Generators | |
Milewski et al. | Comparison of the Ability of Neural Network Model and Humans to Detect a Cloned Voice | |
Anumanchipalli | Intra-lingual and cross-lingual prosody modelling |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A2 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A2 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
122 | Ep: pct application non-entry in european phase | ||
NENP | Non-entry into the national phase in: |
Ref country code: JP |
|
WWW | Wipo information: withdrawn in national office |
Country of ref document: JP |