Web based transcription editor

Abstract

As technology evolves rapidly over the years, the vast majority relies on the Internet to accomplish many daily activities, such as watching videos and TV shows on video sites like YouTube. These videos may include closed captions from a transcript to help different groups of people understand the context better.

However, for most of the time, the transcript is manually prepared by human transcriber(s) who listens to the audio and transcribes the voices into text form and in painstaking detail. The process is very tedious, slow and inefficient. With rapid developments in the area of Speech Recognition, Automatic Speech Recognition (ASR) systems have helped to cut down the manual transcribing work tremendously, with the utilization of deep machine learning and algorithms. However, the ASR output is never error-free due an exhaustive list of factors that can affect the audio quality which the ASR is dependent on. Human intervention is required to review the transcript and make any necessary amendments for quality assurance.

In this project, I will be looking at existing transcribing tools and solutions, analysing their advantages and disadvantages, and explore different technologies that can be integrated into my proposed solution to streamline the process of editing transcripts.