Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method and System for Displaying Personalized Online Video Content based on Domain Ontology and Key-Frame Detection

IP.com Disclosure Number: IPCOM000229341D
Publication Date: 2013-Jul-23
Document File: 2 page(s) / 35K

Publishing Venue

The IP.com Prior Art Database

Related People

Wei Li: INVENTOR [+3]

Abstract

A method and system for displaying personalized online video content based on domain ontology and key-frame detection is disclosed. Personalized video content refers to representing the video as a collection of representative static images which allows a user to quickly survey the video content without consuming the entire video based on preferences of the user. The method and system defines the user preferences based on the domain ontology. The representative images are generated by using key-frame detection algorithms. The key-frame detection algorithms include key-frame extraction and concept detection.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 53% of the total text.

Method and System for Displaying Personalized Online Video Content based on Domain Ontology and Key-Frame Detection

Abstract

A method and system for displaying personalized online video content based on domain ontology and key-frame detection is disclosed.  Personalized video content refers to representing the video as a collection of representative static images which allows a user to quickly survey the video content without consuming the entire video based on preferences of the user.  The method and system defines the user preferences based on the domain ontology.  The representative images are generated by using key-frame detection algorithms.  The key-frame detection algorithms include key-frame extraction and concept detection.

Description

Traditionally, online video sites provide entire videos, which hinders users from viewing a particular video segment of interest.

Disclosed is a method and system for displaying personalized online video content based on domain ontology and key-frame detection.  Personalized video content refers to representing the video as a collection of representative static images which allow a user to quickly survey the video content without consuming the entire video based on preferences of the user.  The preferences of the user are predefined and are domain-specific semantic concepts.  User preferences are considered as the semantic concepts for different types of video content.  For example, the concepts might be goals in a soccer video or a specific movie star name in movie video.  The representative static images are considered as key-frames which are detected by using multi-modal information such as video, audio and text.  The representative images are generated by using key-frame detection algorithms.

 

The method and system defines the user preferences based on the domain ontology.  A large ontology is needed to define a list of concepts of interest and relations between the concepts.  The ontology is constructed from results of formal user studies and data mining of user interaction with online systems such as a query search log.

The method and system generates the representative image sequences using key-frame extraction and concept detection.  A key-frame is a frame which can represent the salient content of a video segment.  A video segment or a video shot ref...