Summary of the paper
| Title | Modified LTSE-VAD Algorithm for Applications Requiring Reduced Silence Frame Misclassification |
|---|---|
| Authors | Iker Luengo, Eva Navas, Igor Odriozola, Ibon Saratxaga, Inmaculada Hernaez, Iñaki Sainz and Daniel Erro |
| Abstract | The LTSE-VAD is one of the best known algorithms for voice activity detection.In this paper we present a modified version of this algorithm, that makes theVAD decision not taking into account account the estimated background noiselevel, but the signal to noise ratio (SNR). This makes the algorithm robust notonly to noise level changes, but also to signal level changes. We compare themodified algorithm with the original one, and with three other standard VADsystems. The results show that the modified version gets the lowest silencemisclassification rate, while maintaining a reasonably low speechmisclassification rate. As a result, this algorithm is more suitable foridentification tasks, such as speaker or emotion recognition, where silencemisclassification can be very harmful. A series of automatic emotionidentification experiments are also carried out, proving that the modifiedversion of the algorithm helps increasing the correct emotion classificationrate. |
| Language | Emotion Recognition/Generation |
| Topics | Tools, systems, applications, Prosody, Emotion Recognition/Generation |
Full paper  |
Modified LTSE-VAD Algorithm for Applications Requiring Reduced Silence Frame Misclassification |
| Bibtex | @InProceedings{LUENGO10.741, author = {Iker Luengo, Eva Navas, Igor Odriozola, Ibon Saratxaga, Inmaculada Hernaez, Iñaki Sainz and Daniel Erro}, title = {Modified LTSE-VAD Algorithm for Applications Requiring Reduced Silence Frame Misclassification}, booktitle = {Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10)}, year = {2010}, month = {may}, date = {19-21}, address = {Valletta, Malta}, editor = {Nicoletta Calzolari (Conference Chair), Khalid Choukri, Bente Maegaard, Joseph Mariani, Jan Odjik, Stelios Piperidis, Mike Rosner, Daniel Tapias}, publisher = {European Language Resources Association (ELRA)}, isbn = {2-9517408-6-7}, language = {english} } |