A comprehensive audio-visual corpus for teaching sound Persian phoneme articulation

Building an audio-visual data corpus is one significant step in audio-visual research. One of the most challenging tasks in computer science is computer-aided speech therapy and language learning. Developing computer applications for training and rehabilitation of the handicapped and helping the hearing and speaking-impaired by facial speech synthesis are among the most helpful, state-of-the-art roles of computer technology in today´s human-machine interacting systems. To date, there have been no audio-visual corpora in Persian language, in that it makes it difficult or even impossible for researchers to carry out studies in the area. This paper gives an indication of the collected Persian audio-visual data corpus. AVA is a comprehensive, systematic collection of both continuous speech and isolated spoken utterances in Persian language. The goal of this project is to facilitate audio-visual research in the language through this data corpus which is available upon request.