H133E080006 Bakke (PI) 9/1/08 to 8/31/13 US Department of Education/NIDRR Rehabilitation Engineering Research Center on Hearing Enhancement The major goal of this project is to develop a comprehensive center addressing needs of individuals with hearing impairment, their families, service providers, hearing aid and cochlear implant manufacturers, and researchers. SRL support on this grant will lead to development of text to speech synthesis and utterance verification software for aural rehabilitation as well as software for aural habilitation of pediatric CI recipients. Role: Project PI for 3 projects
IIS1116799 Patel (PI) 10 /1/11 to 9/30/13 NSF/CSIE HCC: Small: Modeling Acoustic and Articulatory Features for Hybrid Synthesis In this project, we will use conjoint acoustic and articulatory features to model fluent speech for a parametric text to speech synthesis system. Models trained on normally articulating children and adults will be mapped to vocal tract characteristics of dysfluent talkers, allowing us to synthesize fluent speech that sounds like it was produced by the dysfluent talkerÕs vocal tract. Role: Senior Technical Consultant & Subaward PI
Cluster Hassink (PI) 7/1/11 to 6/30/13 Nemours Internal grant award to study obesity prevention and weight management. Bioinformatics component will develop web and messaging services to support closer monitoring and more frequent care provider interaction with patients between clinic visits. Data elements captured within the interactive system will pull to Epic EMR system for ready access by primary care providers. Role: Co-investigator.
Unnumbered Morlet (PI) 6/1/09 to 12/31/10 FriedreichÕs Ataxia Research Alliance Auditory, Speech and Vestibular Abilities in Friedreich Ataxia Describe the hearing and speech deficits associated with Friedreich Ataxia in children and adults. Role: Co-Investigator
IIS0712821 Patel (PI) 7 /1/07 to 6/30/11 (no cost extension) NSF/CSIE Adapting a text to speech synthesizer to convey user identity Novel techniques are proposed to alter Concatenative TTS voices so they sound like other individuals. This would extend the capability of Concatenative TTS systems to individuals who are not able to record enough of their own speech to create a concatenative voice from scratch. Role: Consultant
Unnumbered London (PI) 3/1/10 to 2/28/11 Delaware Health Science Alliance Linking Genotype to Phenotype: A pilot project to create a research data warehouse of biospecimen nd omic information. Implementation of software and hardware to show feasibility of inter-institutional database structures, identify barriers to data sharing, and test approaches to data deidentification .
R42DC006193 Bunnell (PI) 6/1/06 to 5/31/10 NIH/NIDCD Personalized Synthetic Speech using ModelTalker: Development and Evaluation Type: Phase II STTR. Technology transfer grant to make further improvements to the ModelTalker synthesizer and InvTool voice recording program. This phase of the project will focus on improvements to the voice capture and generation stages including changes to the speech coding schemes presently used by ModelTalker. It will also involve development of a Windows CE compatible version of the synthesizer which can run on new AAC devices based on that operating system. Role: PI
R43DC008212 Pal (PI) 9/1/08 to 6/30/09 NIH/NIDCD Use of speech recognition technology for functional hearing assessment Apply Automatic speech recognition to administration and scoring of HINT test. Role: Co-Investigator
R21DC007466 Bunnell (PI) 4/1/05 to 3/31/07 (extension to 3/31/08). NIH/NIDCD Phenotypes and familiality in speech disorders A database of speech production, perception, and acoustic measures will be collected from 250 children along with family history data. HMM based acoustic analysis techniques will be used to characterize children in terms of acoustic phenotypes. These data will be compared to family history data. Role: PI
04/01/04 to 03/31/06 Nemours Research Programs Phonetic acquisition patterns in children with phonological disorders Detailed analyses of utterances recorded from children in the process of acquiring the /r/ - /w/ distinction in production. Will be followed by a prospective study to verify conclusions from the analysis of existing data. Role: Co-Investigator
R41DC006193 Bunnell (PI) 05/15/03 to 4/30/05 NIH Personalized Synthetic Speech Using ModelTalker: Development and Evaluation Phase I STTR project to demonstrate feasibility to commercialize the ModelTalker speech synthesis system for use in AAC devices. Role: PI