2nd speak workshop speech generation in multimodal
play

2nd SPEAK! workshop: Speech Generation in Multimodal Information - PowerPoint PPT Presentation

2nd SPEAK! workshop: Speech Generation in Multimodal Information Systems and Practical Applications Speech synthesis in the Intelligent Personal Communication Support System (IPCSS) Tom Pfeifer Technical University of Berlin e-mail:


  1. 2nd SPEAK! workshop: Speech Generation in Multimodal Information Systems and Practical Applications Speech synthesis in the Intelligent Personal Communication Support System (IPCSS) Tom Pfeifer Technical University of Berlin e-mail: pfeifer@fokus.gmd.de

  2. bitmap format header processing adaption management fax engine in page layouter text text bitmap (email) out bitmap control may include font or postscript interpretetion fax cleaning, formatting mail engine format adaption management in OCR bitmap text fax further conversion (e.g. TTS) out text control Text-to-Fax and Fax-to-Text conversion

  3. . product manufacturer operating hardware supplemental languages system platform hardware TrueTalk Entropic UNIX Sun, SGI E, VisualVoice Stylus Innovations DOS/WIN PC sound card pos- E, sible EASE Expert Systems Dialogic E, required TrueVoice Centigram UNIX, DOS Sun, PC E, Dialogic TTS Dialogic DOS PC Dialogic E, required DecTalk Digital E, Lernout & Lernout & Hous- OS/2, DOS/ various E, Ger, F, NL, Houspie pie WIN, etc. Esp,... Rhetorex TTS Rhetorex UNIX, OS/ PC Rhetorex E, 2, NT required BestSpeech Berkeley Speech- DOS, OS/2, PC etc. Dialogic, Rheto- E, Ger, F, I, J, Technologies UNIX rex, etc. possible NL, Rus Infovox Telia Promotor DOS PC Infovox 500 Ger Infovox Elan Informa- ELAN Informa- DOS PC Televox Psola- Ger, F tique tique 8m Table 1: Evaluation of available TTS systems

  4. generation of perceptible information: perception: human media channels human media channels, technical systems technical technical (examples) representation representation conversion auditory: ear control data control data written speech language sound, music audio (m, n, c, t) audio (m, n, c, t) (natural, technical) midi visual: eye midi movie spoken video (m, n, c, t) video (m, n, c, t) picture language photograph (m,n,c) photograph (m, n, c) graphic video legible text bitmap image bitmap image camera (gif, tiff, fax, ...) (gif, tiff, fax, ...) tactile: skin movie vector image vector image Braille archive page description page description vibration signal (postscript, (postscript, tactile image drawings adobe acrobat) adobe acrobat) vestibular: ear text text photo balance camera numeric numeric haptic: skin handwriting handwriting sensors for grasp force any physical any digital any digital pressure parameter representation representation (temperature, kinaesthetic: body pressure, composed composed force, movement document velocity, document humidity, composed thermic: skin composed voltage, mail mail olfactive: nose smell gustative: tongue taste parameters: m, n: media dependent parameters (frame/sampling rate, quantization, resolution, size, color depth, etc.) c: applied compression technique t: time, duration, etc. Generic Conversion Matrix

  5. � � � PCSS Personal Communications Support System • platform enhancing/ personalizing telecommunications in customer premises networks (CPNs) or of fi ce environements • based on Telecommunications Management Networks (TMN) • addresses two major issues in personal communications: � personal mobility � personalization of services (‘service mobility’) � realization of a personalized communications environment that virtually moves with the user

  6. PCSS Call Processing service generic precessing 1 st Mapping 2 nd Mapping 3 rd Mapping 4 th Mapping Incoming Person to person Person to location Location to virtual virtual communication Call (e.g. Call (processing of communication endpoint (VAP) Call Accept Forwarding) registration data) endpoint (VAP) to Terminal (SAP) to: (dynamic selection of (signal terminal/service) ‘off hook’) Fax R507 call logic dynamic user VAP get comm. evaluation locating selection capabilities SAP Call Handling Address Resolution

  7. PCSS Infrastructure PCSS Applications Communication User Profile Paging Assistance Management Service Service MM Mail Service Manual User User Location (MMM) Registration Information Service Service MM Collabor. Service (MMC) User Presen- Electronic Location PBX / ISDN Server Enabling Client-API Techno- logies User PCSS Profiles Electronic Location IDMIS Techniques Legend: X.700 X.500 MIBs (e.g. Active DIT PCSS Personal Communications Support Badges / System IrSensor Net- DUA Directory User Agent work) IDMIS Inter-Domain Management Core PCSS Information Service

  8. PCSS Generic Service User Profile (X.500) Object Type Must Contain Attributes (some) May Contain Attributes pcssGenServiceUser- pcssPersonalIDKey, pcssAuthenticationNumber, pcssProvidedServiceID, pcssPersonalNumPlanID , Pro fi le (inherited:) pcssPersonalScheduleID , CN, surname pcssAutomaticRegisterID , pcssManualRegisterID , pcssPersonalCallLogicID, (inherited:) description, seeAlso, telephoneNumber, userPassword, userid, textEncodedORAddress, rfc822Mailbox, roomNumber, userClass, homePhone, homePostalAddress, secretary, personalTitle, preferredDeliveryMethod, businessCategory, otherMailbox, mobileTelephoneNumber, pagerTelephoneNumber, organizationalStatus, mailPreferenceOption, personalSignature

  9. PCSS Generic Non General User Data General User Data Service-specific Service User Profile PID User Info Authentication User Data User Personalization Data Personalization Data Profile PNP CRA C. Alerts Diary Personal. Servises Registration Data Registration ManuallyTo AutomaticallyTo ScheduledTo POTS-Profile Call callLogic MMMS-Profile MMCS Profile Mgt. Set of Rules Data MMCS-specific Service-specific User Profile Data Conditions Actions Extensions Legend: containment relationship object instance object group attribute(s) (sub-tree) Manual. Sched.

  10. PCSS Platform PCSS-specific supported User supported Management telecommuni- supported Registration telecommuni- Services & cation services telecommuni- Mgt. Service cation services supported tele- cation services User Communication services Profile Mgt Assistant Service Service PCSS User Agent-API op.xyz() PID_to_SAP() PID_to_ZoneID() (provided as PCSS- PCSS - Applications Framework specific MSCs / MFs) IDMIS (integrated X.500 / X.700 access) PCSS-MIB global PCSS data(X.500) Managed System CN=SOL-MS System Generic PCSS SystemID=sol SAP VAP Service Zone Profiles Profiles User . . . Profiles ServiceProv. . . . = . . Profiles ServerID=ELSl

  11. � � � � Future Perspectives � Interworking between different types of teleservices: Inter-working Function (IWFs) � Dynamic terminal-selection based on Trader function: Virtual Terminal, Virtual Access Points (VAPs) � Interworking of remote PCSSs: Federation of PCSSs � Generic Support for Session Mobility

Recommend


More recommend