Research Project P11:
YouTube Video Classification

Research Project Goal

Given the top XXX most recently posted videos of youtube.com determine which ones are animated, which have excerpts from news broadcasts, which have male vs female protagonists (pick at least two types of categories you would like to recognize).

Can you automatically partition all the top K posted videos into relatively small set of topics/classes ?

Sample Data

All on youtube.com

Tasks

• dowload the videos, propose a sparse representation of the video.
• What are the characteristics of an animated vs real video (compute some simple motion/appearance cues)
• Can you find reliably which videos containt text ?
• Run a OpenCV face detector and determine which videos have faces and which not then partition them to female/male.

Research Project Status

student names here

Point of Contact

Jana Kosecka

Midterm Report

not yet submitted

Final Report

not yet submitted