F1O -Technical Session V:Di-eital Wireless Multimedia Noise Reduction andSpeech Enhancement forHigh-Quality Wireless Handsets

Inwireless communications, thenoise isamoreserious problem than inlandline communications. Toend users, thevoice from awireless handset includes various kinds ofnoises involved fromdifferent sources andcommunication stages, such astheenvironmental background noise, electronic device noise, coding andcomputation noise, communication channel noise, etc.Howtoreduce thenoises andenhance the speech isan important research topic tohigh-quality wireless communications, andnoise reduction is becoming anew feature andselling point towireless handsets. Supported bytheU.S.Government research funding, wehavebeenstudying theproblem foryears and havedeveloped new algorithms andchips fornoise reduction andspeech enhancement forcommercial wireless environments. Ouralgorithm isbased onthemathematic models ofthehumanhearing system, and it canimprove theSNRfor morethan15dBintheenvironments ofcommercial wireless applications. The chip isalowcostandlowpowerconsumption implementation. Itcanbeinserted into wireless handsets or equipments directly. Inthis presentation, wewill demonstrate ournoise reduction algorithms with various background noises incomparison with other products onthemarket. Ourchips withpre-loaded algorithm areready forcell phone manufactures for newhigh-quality wireless handsets. BIOGRAPHY Dr.Qi(Peter) Lireceived hisPh.D. degree inelectrical engineering fromtheUniversity ofRhodeIsland, Kingston, in1995. From1988to1994, heworked atF.M.Engineering andResearch, Norwood, MA, where heengaged inpattern recognition, real-time systems, andnetworks. In1991, heattended Harvard University tostudy statistical theory andmethods. From1995to2002, hewaswithBellLaboratories, Lucent Technologies, Murray Hill, NJ, asaMemberofTechnical Staff intheMultimedia Communications Research Lab, where hisresearch wasinspeech andspeaker recognition, biometric authentication, frontendsignal processing, speech modeling, andmultimedia systems; hisresearch results wereimplemented in Lucent products. In2002, Dr.Liestablished LiCreative Technologies (LcT), Inc., located inFlorham Park, NJ.LcTisa high-tech company inR&D forsignal processing, multimedia applications, andhomeland security. Since 2003, LcThasbeenawarded several research contracts byDARPA,NAVY,andMDA oftheDepartment ofDefense. Thecompanyisconducting research inacoustic-signal processing forhandheld devices, emotion statedetection, speaker recognition, newfeatures forspeech processing, data-driven prognostics, decision models andsystems, etc.Thecompany isalso developing products based onits newresearch results, andisinterested inlicensing newtechniques developed inthecompany fortheindustry. Dr.Liisasenior memberofIEEE.Hehasbeenactive asareviewer forseveral journals, including IEEE Transactions onSpeech andAudio Processing. Hehasalso beenaLocal Chair fortheIEEEWorkshop on Automatic Identification andacommittee memberforseveral IEEEconferences andworkshops. Hehas received abest-paper award, anachievement award, several Bell Labspatent awards, and2004NewJersey Small Business Success Award. Dr.Liislisted inWho's WhoinAmerica (Millennium and2001Editions) andinWho'sWhoinExecutives andProfessionals (2004 Edition).