other,19-2-H93-1076,ak processors </term> but operate directly on <term> audio and scanned image data </term> . A second , related theme is the
tech,24-1-H93-1076,ak <term> recognition technology </term> in <term> document-oriented applications </term> . One is the development of <term>
other,22-3-H93-1076,ak user-specified information from documents with <term> signal content </term> . This paper discusses three research
tech,13-5-H93-1076,ak in the research that enable novel <term> signal-based document processing functionality </term> . We present an implemented <term>
tech,14-4-H93-1076,ak PARC that exemplify these themes : a <term> text-image editor </term> \ [ 1 \ ] , a <term> wordspotter </term>
tech,25-4-H93-1076,ak \ ] , a <term> wordspotter </term> for <term> voice editing and indexing </term> \ [ 12 \ ] , and a <term> decoding
tech,40-4-H93-1076,ak <term> decoding framework </term> for <term> scanned-document content retrieval </term> \ [ 4 \ ] . The discussion focuses
tech,5-1-H93-1076,ak </term> . Two themes have evolved in <term> speech and text image processing work </term> at Xerox PARC that expand and redefine
tech,13-2-H93-1076,ak provide functionality similar to that of <term> text processors </term> but operate directly on <term> audio
tech,37-4-H93-1076,ak indexing </term> \ [ 12 \ ] , and a <term> decoding framework </term> for <term> scanned-document content
tech,23-4-H93-1076,ak text-image editor </term> \ [ 1 \ ] , a <term> wordspotter </term> for <term> voice editing and indexing
tech,21-1-H93-1076,ak that expand and redefine the role of <term> recognition technology </term> in <term> document-oriented applications
tech,5-2-H93-1076,ak </term> . One is the development of <term> systems </term> that provide functionality similar
tech,9-3-H93-1076,ak second , related theme is the use of <term> speech and text-image recognition </term> to retrieve arbitrary , user-specified
hide detail