A text-to-speech system (TTS) is one of the human-machine interfaces using speech. TTSs, which can be implemented in software or hardware, convert normal language text into speech. TTSs are implemented in many applications such as car navigation systems, information retrieval over the telephone, voice mail, speech-to-speech translation systems, and comparable ones with a goal of synthesizing speech with natural human voice characteristics. Modem text to speech systems provide users access to multitude of services integrated in interactive voice response systems. Telephone customer service is one of the examples of rapidly proliferating text to speech functionality in interactive voice response systems.
Many systems employing a TTS engine require human-like voice output to speak static content (prompts). When the recording person is not available, a prompt generation tool is usually used to help generate such prompts. A prompt generation tool helps people to manipulate text-to-speech output to achieve better prosody, naturalness, etc. A common deficiency of these tools is the lack of ease of use and efficiency to get a satisfying result, because the representation of waveforms is hard to be understood by people with little or no speech synthesis background.
This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This summary is not intended to exclusively identify key features or essential features of the claimed subject matter, nor is it intended as an aid in determining the scope of the claimed subject matter.
Embodiments are directed to an interactive prompt generation and TTS optimization tool with an easy-to-understand graphical user interface representation of the TTS process that can be employed to guide user through different speech recognition and synthesis technologies for the generation of initial text-to-speech output.
These and other features and advantages will be apparent from a reading of the following detailed description and a review of the associated drawings. It is to be understood that both the foregoing general description and the following detailed description are explanatory and do not restrict aspects as claimed.
As briefly described above, an interactive prompt generation and TTS optimization tool with an easy-to-understand graphical user interface representation of the TTS process may be employed to guide users through different speech recognition and synthesis technologies for the generation of initial text-to-speech output. These aspects may be combined, other aspects may be utilized, and structural changes may be made without departing from the spirit or scope of the present disclosure. The following detailed description is therefore not to be taken in a limiting sense, and the scope of the present invention is defined by the appended claims and their equivalents.
While the embodiments will be described in the general context of program modules that execute in conjunction with an application program that runs on an operating system on a personal computer, those skilled in the art will recognize that aspects may also be implemented in combination with other program modules.
Generally, program modules include routines, programs, components, data structures, and other types of structures that perform particular tasks or implement particular abstract data types. Moreover, those skilled in the art will appreciate that embodiments may be practiced with other computer system configurations, including hand-held devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, minicomputers, mainframe computers, and comparable computing devices. Embodiments may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.
Embodiments may be implemented as a computer-implemented process (method), a computing system, or as an article of manufacture, such as a computer program product or computer readable media. The computer program product may be a computer storage medium readable by a computer system and encoding a computer program that comprises instructions for causing a computer or computing system to perform example process(es). The computer-readable storage medium can for example be implemented via one or more of a volatile computer memory, a non-volatile memory, a hard drive, a flash drive, a floppy disk, or a compact disk, and comparable media.
Throughout this specification, the term “TTS” is a Text To Speech system. TTS system refers to a combination of software and hardware components for converting text to speech. Examples of platforms include, but are not limited to, an Interactive Voice Response (IVR) system such as those used in telephone, vehicle applications, and similar implementations. The term “server” generally refers to a computing device executing one or more software programs typically in a networked environment. However, a server may also be implemented as a virtual server (software programs) executed on one or more computing devices viewed as a server on the network. More detail on these technologies and example operations is provided below. Also, the term “engine” is used to refer to a self contained software application that has input(s) and an output(s).
Text to speech system (TTS) 112 converts text 102 to speech 110 by performing an analysis on the text to be converted, an optional linguistic analysis, and a synthesis putting together the elements of the final product speech. The text to be converted may be analyzed by text analysis component 104 resulting in individual words, which are analyzed by the linguistic analysis component 106 resulting in phonemes. Waveform generation component 108 synthesizes output speech 110 based on the phonemes.
Depending on a type of TTS, the system may include additional components. The components may perform additional or fewer tasks and some of the tasks may be distributed among the components differently. For example, text normalization, pre-processing, or tokenization may be performed on the text as part of the analysis. Phonetic transcriptions are then assigned to each word, and the text divided and marked into prosodic units, like phrases, clauses, and sentences. This text-to-phoneme or grapheme-to-phoneme conversion is performed by the linguistic analysis component 106.
Two major types of generating synthetic speech waveforms are concatenative synthesis and formant synthesis. Concatenative synthesis is based on the concatenation (or stringing together) of segments of recorded speech. While producing close to natural-sounding synthesized speech, in this form of speech generation differences between natural variations in speech and the nature of the automated techniques for segmenting the waveforms may sometimes result in audible glitches in the output. Sub-types of concatenative synthesis include unit selection synthesis, which uses large databases of recorded speech. During database creation, each recorded utterance is segmented into some or all of individual phones, diphones, half-phones, syllables, morphemes, words, phrases, and sentences. An index of the units in the speech database is then created based on the segmentation and acoustic parameters like the fundamental frequency (pitch), duration, position in the syllable, and neighboring phones. At runtime, the desired target utterance is created by determining the best chain of candidate units from the database (unit selection).
Another sub-type of concatenative synthesis is diphone synthesis, which uses a minimal speech database containing all the diphones (sound-to-sound transitions) occurring in a language. A number of diphones depends on the phonotactics of the language. At runtime, the target prosody of a sentence is superimposed on these minimal units by means of digital signal processing techniques such as linear predictive coding. Yet another sub-type of concatenative synthesis is domain-specific synthesis, which concatenates prerecorded words and phrases to create complete utterances. This type is more compatible for applications where the variety of texts to be outputted by the system is limited to a particular domain.
In contrast to concatenative synthesis, formant synthesis does not use human speech samples at runtime. Instead, the synthesized speech output is created using an acoustic model. Parameters such as fundamental frequency, voicing, and noise levels are varied over time to create a waveform of artificial speech. While the speech generated by formant synthesis may not be as natural as one created by concatenative synthesis, formant-synthesized speech can be reliably intelligible, even at very high speeds, avoiding the acoustic glitches that are commonly found in concatenative systems. High-speed synthesized speech is, for example, used by the visually impaired to quickly navigate computers using a screen reader. Formant synthesizers can be implemented as smaller software programs and can, therefore, be used in embedded systems, where memory and microprocessor power are especially limited.
HMM-based speech synthesis is also an acoustic model based synthesis method employing Hidden Markov Models. Frequency spectrum (vocal tract), fundamental frequency (vocal source), and duration (prosody) of speech are commonly modeled simultaneously by HMMs. Speech waveforms are then generated from HMMs themselves based on a maximum likelihood criterion.
HMM based text to speech systems (HTSs), which can be automatically trained, can generate natural and high quality synthetic speech and reproduce voice characteristics of the original speaker. HTSs utilize the flexibility of HMMs such as context-dependent modeling, dynamic feature parameters, mixture of Gaussian densities, tying mechanism, speaker and environment adaptation techniques.
There are many parameters in speech synthesis, variation of which may result in different perception by different users. For example, pitch, dialect, gender of speaker, and so on may influence how synthesized speech is perceived by users. In conventional prompt generation tools, an initial wave is first generated by the core text-to-speech engine. Then, the user needs to adjust acoustic features like duration, prosody, energy, etc. on top of this wave. Due to the limitation of current concatenation text-to-speech technology, the voice quality gap between the initial synthesized voice and the expected result is usually significant because of poor prosody prediction algorithms and similar challenges.
The system illustrated in diagram 200 includes a TTS engine core 214 and the interactive prompt generation/TTS optimization tool 220. As discussed above, the TTS engine core 214 may receive prosody information extracted from an HTS system 216 or real prosody information extracted from the user's own voice 218. TTS engine core 214 provides information for generating the initial waveform to wave synthesizer 224 of the interactive tool 220. Wave synthesizer 224 may also receive text input from the user in form of prompt script (212).
Interactive tool 220 enables an iterative process of checking the quality (226) of the synthesized wave with feedback (222) provided to wave synthesizer 224. Feedback process (222) may include correction of frontend errors, acoustic unit reselection, unit reselection with prosody adjustment, and similar modifications. Once the quality is deemed acceptable, the end product may be saved (232) in a structured project file (e.g. an xml file) 234, as a recording (236), and as binary data (voice font 238) for use by the TTS engine core. The recordings may also be provided to a prompt engine 240 for further processing depending on the application type.
The prosody information (pitch/duration/energy) may be abstracted to simple visual presentation in an interactive tool according to embodiments. For example, the pitch curve may be displayed in a simple format and duration/energy represented through width and/or color of GUI elements. These representations may reduce a user's learning curve and operation complexity without losing tuning ability.
Since HTS technology can generate better prosody information, prosody information is extracted from an HTS system and used to guide the concatenative TTS system in a system according to one embodiment. This helps the system to generate better initial waveforms. When the initial synthesized voice is closer to an expected result, the users' tuning effort is simplified increasing an efficiency of the TTS system.
Furthermore, the user is also enabled to speak the desired output for recording by the tool. The interactive tool extracts key acoustic information including pitch variation, duration, and energy of each phoneme to guide the text-to-speech engine in generating the initial synthesized voice. With such guidance, the users' tuning effort is again significantly simplified. Users with little or no speech prosody knowledge may be enabled to utilize the interactive tool to adjust prosody information.
When a new session is selected a new window 346 may be opened enabling the user to specify a name for the session, a location for saving the session, and a prompt type. Moreover, the user may be enabled to input the text to be converted to speech for the new prompt.
Unit sequence 456 displays acoustic units comprising the prompt presented in graphical format such that the user can visually determine a pitch and length for each unit. For example, an incline in the graphical representation may indicate higher pitch, while the opposite indicates a lower pitch. Upon receiving a selection of a word in the word sequence from the user, the user interface may also display a link between the selected word and corresponding acoustic units.
As individual phonetic characters are selected information associated with them (562) such as usage of the character in an example word and an audio playback of the same may be provided to the user. The phonetic character list may be modified depending on which phonetic alphabet is used.
Additional elements 676 of the user interface shown in diagram 600 include a click-on button for modifying a pitch of a currently selected acoustic unit, a slide scale for adjusting the duration of the currently selected acoustic unit, and a second slide scale for adjusting an energy of the currently selected acoustic unit. A click-on button for playing back the current selection is also provided as part of the user interface.
The user interface, which may be a activated as a new window is identified as “Pitch Patterns for m+ay” (778), where m+ay is the currently selected acoustic unit. Various pitch patterns 780 are provided in a visual form for the user to select. The visual format is a user-friendly way of enabling the user to modify the pitch of the prompt on a unit by unit basis without having to guess how each alternative sounds. A preview button 782 enables the user to listen to alternative pitch patterns.
Differently from the user interface of
The “Manage Sessions” user interface shown in diagram 900 is for enabling the user to efficiently find, open, save, and close sessions for various prompts. A storage location 990 may be provided, where the user can type in the location or browse through a different method. Found sessions may be listed (992) with their assigned names and the corresponding text or each session.
A playback button enables the user to listen to selected prompts without activating an edit user interface. An open session button enables the user to activate the edit user interface where he/she can edit various aspects the synthesized prompt as discussed previously.
The TTS based systems, components, configurations, user interface elements, and mechanisms illustrated above are for example purposes and do not constitute a limitation on embodiments. An interactive TTS optimization tool according to embodiments may be implemented with other components and configurations using the principles described herein.
As discussed previously, client devices 1011-1014 are used to facilitate communications employing a variety of modes between users of the TTS system. TTS related information such as pronunciation elements, training data, and the like may be stored in one or more data stores (e.g. data store 1019), which may be managed by any one of the servers 1016 or by database server 1018.
Network(s) 1010 may comprise any topology of servers, clients, Internet service providers, and communication media. A system according to embodiments may have a static or dynamic topology. Network(s) 1010 may include a secure network such as an enterprise network, an unsecure network such as a wireless open network, or the Internet. Network(s) 1010 may also coordinate communication over other networks such as PSTN or cellular networks. Network(s) 1010 provides communication between the nodes described herein. By way of example, and not limitation, network(s) 1010 may include wireless media such as acoustic, RF, infrared and other wireless media.
Many other configurations of computing devices, applications, data sources, and data distribution systems may be employed to implement an interactive TTS optimization and prompt generation tool. Furthermore, the networked environments discussed in
TTS application 1122 may be any application that synthesizes speech as discussed previously. Interactive tool 1124 may be an integral part of TTS application 1122 or a separate application. Interactive tool 1124 may enable users to provide text for conversion to speech, visually and audibly provide feedback on alternatives for different pronunciation of the text, and enable the user to modify the parameters. This basic configuration is illustrated in
Computing device 1100 may have additional features or functionality. For example, the computing device 1100 may also include additional data storage devices (removable and/or non-removable) such as, for example, magnetic disks, optical disks, or tape. Such additional storage is illustrated in
Computing device 1100 may also contain communication connections 1116 that allow the device to communicate with other devices 1118, such as over a wireless network in a distributed computing environment, a satellite link, a cellular link, and comparable mechanisms. Other devices 1118 may include computer device(s) that execute communication applications, other directory or presence servers, and comparable devices. Communication connection(s) 1116 is one example of communication media. Communication media can include therein computer readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and includes any information delivery media. The term “modulated data signal” means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared and other wireless media.
Example embodiments also include methods. These methods can be implemented in any number of ways, including the structures described in this document. One such way is by machine operations, of devices of the type described in this document.
Another optional way is for one or more of the individual operations of the methods to be performed in conjunction with one or more human operators performing some. These human operators need not be collocated with each other, but each can be only with a machine that performs a portion of the program.
Process 1200 begins with optional operation 1210, where text to be converted to speech is received from a user. The interactive tool may enable the user to provide the text by typing, by importing from a document, or any other method. At operation 1220, a first pass at synthesis is made employing default (and/or user preferred) parameters. The synthesized speech is provided in audible form to the user with a playback option, and a phonetic breakdown of the prompt is also presented.
At operation 1230, the user is enabled to modify various parameters of the TTS system based on visual cues provided by the interactive tool as discussed in the example user interfaces previously. As part of the modification process, alternative pronunciations may be provided visually and audibly (playback option).
When the user is finished and indicates that he/she would like to save the end product, the modified prompt is saved at operation 1240 for subsequent use by another application such as an Interactive Voice Response (IVR) system.
The operations included in process 1200 are for illustration purposes. An interactive TTS optimization and prompt generation tool may be implemented by similar processes with fewer or additional steps, as well as in different order of operations using the principles described herein.
The above specification, examples and data provide a complete description of the manufacture and use of the composition of the embodiments. Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims and embodiments.
Number | Name | Date | Kind |
---|---|---|---|
7881934 | Endo et al. | Feb 2011 | B2 |
20050060158 | Endo et al. | Mar 2005 | A1 |
20080167875 | Bakis et al. | Jul 2008 | A1 |
20090228271 | Desimone | Sep 2009 | A1 |
Number | Date | Country | |
---|---|---|---|
20100312565 A1 | Dec 2010 | US |