Automatically Generating Natural Language Descriptions for Videos

April 7, 2016 @ 6:00 pm – 7:00 pm America/New York Timezone
City: Rochester

Co-sponsored by: AdvanceRIT

Title: Automatically Generating Natural Language Descriptions for Videos

Presentor: Subhashini Venugopalan, UT Austin

For most people, watching a brief video and describing what happened (in words) is an easy task. For machines, extracting the meaning from video pixels and generating a sentence description or a caption is a very complex problem. In this talk I will present some of my work on developing models that can automatically generate natural language descriptions for events in videos. These models integrate recent advances in computer vision, natural language processing, and “deep” machine learning to automatically describe short video clips . I will also show how these models perform on clips from Youtube and movie snippets.


Subhashini Venugopalan is a PhD candidate in the Computer Science department at the University of Texas at Austin. Her research focuses on deep learning techniques to automatically generate descriptions of events in videos.  She is advised by Prof. Raymond Mooney. Subhashini holds a Masters degree in Computer Science from Indian Institute of Technology, Madras and a Bachelors degree in Information Technology from National Institute of Technology, Karnataka, India. Subhashini also has experience as a Software Engineer (intern) at Google and IBM, Research.

Speaker(s): Subhashini Venugopalan ,


Presentation 6-6:45

Q/A 6:45-7pm

Bldg: Rochester Institute of Technology, INS bldg, Room 1140
Rochester, New York