Virginia's George Mason University hopes to one day make captioning in the classroom completely automated. A caption proposal has been approved at the school that would allow faculty members to upload their audio files to a server. From there, automated voice-recognition software called Docsoft:AV produces a fully captioned transcript of the classroom lecture. Student workers will review what the program produces until the automation is accurate enough to work on its own. It must overcome poor audio quality, the different accents of speakers, and other variables. George Mason has a number of veterans with hearing impairments who are are taking classes and using the CART system in the classroom (CART stands for Communication Access Realtime Translation). This real-time text translation is expensive and requires someone to transcribe lecture from another location.