Sixth International Conference on Spoken Language Processing
Transcoding refers to the process of converting HTML files in order to adapt the physical attributes of each device. We call our transcoding "Semantic Transcoding" because it transcodes at the deep semantic level. The proxy server consists of translation, text summarization, video summarization, text-to-speech, speech recognition, and so on. External annotation is also possible. Semantic transcoding enables the visually challenged to browse Web pages by changing text to speech, and to understand the pictures by reading external annotated comments. The hearing impaired are provided with superimposed dialog generated by the recognition of speech and scene. Furthermore, we will propose a rapid "diagonal" reading method by listening to speech converted from text on the basis of linguistic information such as syntax, new and old information for the blind and visually impaired in the internet speech browser realized by semantic transcoding.
Bibliographic reference. Torihara, Shinichi / Nagao, Katashi (2000): "Semantic transcoding: making the handicapped and the aged free from their barriers in obtaining information on the web", In ICSLP-2000, vol.3, 1117-1120.