Deep Learning and Online Video: Advances in Transcription, Automated Indexing, and Manipulation

In recent years, the amount of video content created and uploaded to the Internet has grown exponentially. Video content has unique accessibility challenges: indexing, transcribing, and searching video has always been very labor intensive, and there were no automated ways of searching videos for specific content. New software tools that use deep learning methods are automating some of these processes, making video content more discoverable and useful. There are also many new tools for processing and manipulating video in interesting ways. This column will briefly discuss the idea of deep learning and how deep learning tools can be used to transcribe, translate, search, and even manipulate videos. It will suggest ways that librarians can use these tools to help their institutions better manage video content. It also includes a list of video-related software tools.