In some embodiments, a transcription generation technique other than the transcription generation technique initially selected may be selected. For example, in some instances, transcriptions obtained via a particular transcription generation technique may receive unfavorable user ratings such that another transcription generation technique may be selected to obtain transcriptions of subsequent communication sessions.

In some embodiments, a first transcription generation technique may be different from a second transcription generation technique based on how the transcription generation is performed. For example, the first transcription generation technique may include generation of the transcription by a fully machine based automatic speech recognition (ASR) system. Fully machine based ASR systems may operate without human intervention and may be referred to in this disclosure as automatic systems. Alternatively or additionally, the second transcription generation technique may include generation of the transcription by a re-voicing transcription system.

Re-voicing transcription systems, referred to in this disclosure as re-voicing systems, may receive and broadcast audio to a captioning agent (e.g., a human captioning agent). The captioning agent may listen to the broadcast and speak the words from the broadcast. The words spoken by the captioning agent are captured to generate re-voiced audio. The re-voiced audio may be used by a speech recognition program to generate the transcription of the audio. In some embodiments, the speech recognition program may be trained to the voice of the captioning agent.