Voice Recognition Based Automated Teleprompter Application

In the realm of academic or other oral presentations, managing cues and scripts on paper may be cumbersome and conducive to unnecessary complications. Teleprompters may provide a helpful alternative; however, they usually operate at a constant speed, causing presenters to pause, stutter or lose track of their train of thought. The current paper presents the Prophet project, which aims to create an application providing assistance and automation to speakers in practicing and delivering their oral presentations using a voice recognition driven automated teleprompter. The project consists of a mobile application and a document converter service. The former focuses on user need: one can create new texts, import already existing ones and edit them using an in-built rich-text editor. The resulting text can then be used in the automated teleprompter, which uses natural language processing to indicate the progress of the speech. Existing documents may be imported via a proprietary service which extracts the content of text files with various formats, and converts said content into a uniform text format for the application to use.