Modipa, TDavel, MDe Wet, Febe2012-03-062012-03-062010-10Modipa, T, Davel, M and De Wet, F. Acoustic modelling of Sepedi affricates for ASR. 2010 Annual Research Conference of the South African Institute for Computer Scientists and Information Technologists (SAICSIT 2010), Bela Bela, South Africa, 11-13 October 2010978-1-60558-950-3http://delivery.acm.org/10.1145/1900000/1899552/p394-modipa.pdf?ip=146.64.81.8&acc=ACTIVE%20SERVICE&CFID=88545526&CFTOKEN=51832324&__acm__=1331025181_6c0810155717926dddd97de90a1b4bf0http://hdl.handle.net/10204/56312010 Annual Research Conference of the South African Institute for Computer Scientists and Information Technologists (SAICSIT 2010), Bela Bela, South Africa, 11-13 October 2010. This is an ABSTRACT ONLYAutomatic speech recognition (ASR) systems are increasingly being developed for under-resourced languages, especially for use in multilingual spoken dialogue systems. We investigate different approaches to the acoustic modelling of Sepedi affricates for ASR. We determine that it is possible to model various of these complex consonants as a sequence of much simpler sounds. This approach reduces the Sepedi phoneme inventory from 45 to 32, resulting in simpler dictionary development and transcription processes, as well as more accurate acoustic modelling.enArtificial intelligenceNatural language processingSpeech recognition and synthesisSepediAcoustic modelsPhonemesAffricatesConference PresentationModipa, T., Davel, M., & De Wet, F. (2010). Acoustic modelling of Sepedi affricates for ASR. ACM Digital Library. http://hdl.handle.net/10204/5631Modipa, T, M Davel, and Febe De Wet. "Acoustic modelling of Sepedi affricates for ASR." (2010): http://hdl.handle.net/10204/5631Modipa T, Davel M, De Wet F, Acoustic modelling of Sepedi affricates for ASR; ACM Digital Library; 2010. http://hdl.handle.net/10204/5631 .TY - Conference Presentation AU - Modipa, T AU - Davel, M AU - De Wet, Febe AB - Automatic speech recognition (ASR) systems are increasingly being developed for under-resourced languages, especially for use in multilingual spoken dialogue systems. We investigate different approaches to the acoustic modelling of Sepedi affricates for ASR. We determine that it is possible to model various of these complex consonants as a sequence of much simpler sounds. This approach reduces the Sepedi phoneme inventory from 45 to 32, resulting in simpler dictionary development and transcription processes, as well as more accurate acoustic modelling. DA - 2010-10 DB - ResearchSpace DP - CSIR KW - Artificial intelligence KW - Natural language processing KW - Speech recognition and synthesis KW - Sepedi KW - Acoustic models KW - Phonemes KW - Affricates LK - https://researchspace.csir.co.za PY - 2010 SM - 978-1-60558-950-3 T1 - Acoustic modelling of Sepedi affricates for ASR TI - Acoustic modelling of Sepedi affricates for ASR UR - http://hdl.handle.net/10204/5631 ER -