Post Facto

A Blog by Holly French


I spent the summer of 2011 at Carleton College doing research under Amy Csizmar Dalal with Jie Lin and Tung Phan.

Our research group focused on identifying application-layer measurements to predict users' subjective ratings of streaming video. We set up a system to collect video stream information from Flash videos streaming from a web browser. We then applied data mining techniques (specifically Dynamic Time Warping) to assign ratings to individual video streams. Using this, we were able to assign ratings with anywhere from 70-90% accuracy.

We have some preliminary results which show that we can accurately predict stream quality ratings with as little as ten seconds of stream data, which indicate that we can work towards predicting stream quality in real time.

The results of our summer research appeared in IPCCC. You can read about it here:
H. French, J. Lin, T. Phan, A. Csizmar Dalal. “Real Time Video QoE Analysis of RTMP Streams.” (Poster). Appeared in Proceedings of the 30th IEEE International Performance Computing and Communications Conference (IPCCC), Orlando, FL, November 2011.