Arabic in my Hand: Small-footprint Synthesis of Egyptian Arabic

摘要

The research described in this paper addresses the dual concerns of synthesis of Arabic, a language that has shot to prominence in the past few years, and synthesis on a handheld device, realization of which presents difficult software engineering problems. Our system was developed in conjunction with the DARPA BABYLON project, and has been integrated with English synthesis, English and Arabic ASR, and machine translation on a single off-the-shelf PDA. We present a concatenative, general-domain Arabic synthesizer that runs 7 times faster than real time with a 9MB footprint. The voice itself was developed over only a few months, without access to costly prepared databases. It has been evaluated using standard test protocols with results comparable to those achieved by English voices of the same size with the same level of development.