Unsupervised Stream-Weights Computation in Classification and Recognition Tasks

E. Sanchez-Soto 1 A. Potamianos 2 Khalid Daoudi 3
1 IRIT - Institut de recherche en informatique de Toulouse
2 E.C.E - Department of Electronic and Computer Engineering [Crete]
3 GeoStat - Geometry and Statistics in acquisition data
Inria Bordeaux - Sud-Ouest
Abstract : In this paper, we provide theoretical results on the problem of optimal stream weight selection for the multi-stream classi- fication problem. It is shown, that in the presence of estimation or modeling errors using stream weights can decrease the total classification error. Stream weight estimates are computed for various conditions. Then we turn our attention to the problem of unsupervised stream weights computation. Based on the theoretical results we propose to use models and “anti-models” (class- specific background models) to estimate stream weights. A non-linear function of the ratio of the inter- to intra-class distance is used for stream weight estimation. The proposed unsupervised stream weight estimation algorithm is evaluated on both artificial data and on the problem of audio-visual speech classification. Finally the proposed algorithm is extended to the problem of audio- visual speech recognition. It is shown that the proposed algorithms achieve results comparable to the supervised minimum-error training approach under most testing conditions.
Document type :
Journal articles
Complete list of metadatas

Cited literature [23 references]  Display  Hide  Download
https://hal.inria.fr/inria-00438846
Contributor : Khalid Daoudi <>
Submitted on : Friday, December 4, 2009 - 7:10:38 PM
Last modification on : Monday, January 13, 2020 - 10:04:04 AM
Long-term archiving on: Thursday, June 17, 2010 - 7:40:54 PM

File

Final-TrSAL09.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : inria-00438846, version 1

Collections

CNRS | INRIA | UNIV-TLSE2 | SMS | UNIV-TLSE3 | UT1-CAPITOLE

Citation

E. Sanchez-Soto, A. Potamianos, Khalid Daoudi. Unsupervised Stream-Weights Computation in Classification and Recognition Tasks. IEEE Transactions on Audio, Speech and Language Processing, Institute of Electrical and Electronics Engineers, 2009, 17 (3), pp.436--445. ⟨inria-00438846⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

358

Files downloads

1206

 

 

Contact

archive-ouverte@inria.fr