| Popis: |
Computer‐based visual speech training aids are potentially useful feedback tools for hearing impaired people. In this paper, a training aid for the articulation of short Consonant‐Vowel‐Consonant (CVC) words is presented using an integrated real‐time display of phonetic content and loudness. Although not yet extensively tested with hearing‐impaired listeners, real‐time signal processing has been developed, flow‐mode displays have been developed, and accuracy tests have been completed with a large database of CVC short words. A combination of Hidden Markov Models and time‐delay neural networks are used as the primary method for the acoustic‐phonetic transformations used in the display. [Work partially supported by NSF grant BES‐9977260.] |
