Announcing AVA: A Finely Labeled Video Dataset for Human Action Understanding Announcing AVA: A Finely Labeled Video Dataset for Human Action UnderstandingSource: Google ResearchPublished on 2017-10-19By Research Blogfeel free to share...Facebook0Google+0Twitter0LinkedinReddit0email