Abstract
This paper describes the system proposed for the Spoken Web Search task at Mediaeval 2012 campaign. We use an audio-only system based on our new called Cumulative Dynamic Time Warping (CDTW) algorithm. This algorithm combines the scores of all the alignment paths and allows for the learning of different cost functions for each kind of step in the alignment matrix (diagonal, horizontal and vertical). The results obtained with basic audio descriptors show the promising potential of our algorithm.
Original language | English |
---|---|
Journal | CEUR Workshop Proceedings |
Volume | 927 |
State | Published - 2012 |
Externally published | Yes |
Event | Multimedia Benchmark Workshop, MediaEval 2012 - Pisa, Italy Duration: 4 Oct 2012 → 5 Oct 2012 |