Video Summarization

The application uses Computer Vision and Neural Networks to automatically create a trailer of a video file. Different on the category and contents of the video file, different algorithms are employed to create the best summary.

For example, if subtitles are available, they will be used to generate the summary. If a subtitle is not available speech-to-text is used to generate the summary. If neither is available, the Neural Network will determine the most salient points in the Video and use those as guidelines to generate the summary. The users can control the duration of the summary created.

 

Problem

Normally when content creators create trailers for marketing purposes, they have to analyze the whole video before creating the trailers. This approach is not scalable when the volume of videos is large and the deadlines are tight.

Solution

A better approach is to automate the process so the application generates the first cut and Editor/s create the final cut. The flexibility of creating summaries of different durations further streamlines the trailer creation workflow.


 

Use cases

Sports

News

Personalized Playlist

Industry

Media and Sports networks

News Channels