×
inauthor:"Michael Alfred Smith" from books.google.com
The goal of this work is to show the utility of integrating language and image understanding techniques for video skimming by extraction of significant information, such as specific objects, audio keywords and relevant video structure.