Video summarization via crowdsourcing
- 7 May 2011
- proceedings article
- Published by Association for Computing Machinery (ACM)
- p. 1531-1536
- https://doi.org/10.1145/1979742.1979803
Abstract
Although video summarization has been studied extensively, existing schemes are neither lightweight nor generalizable to all types of video content. To generate accurate abstractions of all types of video, we propose a framework called Click2SMRY, which leverages the wisdom of the crowd to generate video summaries with a low workload for workers. The framework is lightweight because workers only need to click a dedicated key when they feel that the video being played is reaching a highlight. One unique feature of the framework is that it can generate different abstraction levels of video summaries according to viewers' preferences in real time. The results of experiments conducted to evaluate the framework demonstrate that it can generate satisfactory summaries for different types of video clips.Keywords
This publication has 4 references indexed in Scilit:
- Quantifying QoS requirements of network servicesPublished by Association for Computing Machinery (ACM) ,2011
- Dynamic video summarization using two-level redundancy detectionMultimedia Tools and Applications, 2008
- Video abstraction: A systematic review and classificationACM Transactions on Multimedia Computing, Communications, and Applications, 2007
- Evolving video skims into useful multimedia abstractionsPublished by Association for Computing Machinery (ACM) ,1998