Investigating the Feasibility Factors of Synthetic Sign Language Visualization Methods on Mobile Phones

摘要

The SASL project at the University of the Western Cape is in the process of creating a machine translation system on mobile phones that can automate the translation process between English and South African Sign Language (SASL). This will supplement the use of highly trained interpreters that are expensive and scarce. In order to do this it is necessary to determine whether or not sign language visualization on mobile phones is feasible. This paper describes the research carried out in order to determine the feasibility of four sign language visualization methods on mobile phones. Three factors were considered: the intelligibility; the bandwidth consumption; and the power consumption associated with each method, with the intelligibility being the most important factor. Three of these methods were synthetic sign language visualization methods. It was found that, on average, 65% of signs were recognizable. This figure was affected by sign language dialect variations and increased to 85% with this factor removed. On average, the animations of methods were found to have file sizes of 162 KB and caused a 180% increase in power consumption, compared to the idle operation, on the mobile phone.