Search Results - "Saon, G."

Advances in speech transcription at IBM under the DARPA EARS program by Chen, S.F., Kingsbury, B., Lidia Mangu, Povey, D., Saon, G., Soltau, H., Zweig, G.

“…This paper describes the technical and system building advances made in IBM's speech recognition technology over the course of the Defense Advanced Research…”
Get full text

Advances in Arabic Speech Transcription at IBM Under the DARPA GALE Program by Soltau, H., Saon, G., Kingsbury, B., Kuo, H.-K.J., Mangu, L., Povey, D., Emami, A.

“…This paper describes the Arabic broadcast transcription system fielded by IBM in the GALE Phase 2.5 machine translation evaluation. Key advances include the…”
Get full text

Maximum likelihood discriminant feature spaces by Saon, G., Padmanabhan, M., Gopinath, R., Chen, S.

“…Linear discriminant analysis (LDA) is known to be inappropriate for the case of classes with unequal sample covariances. There has been an interest in…”
Get full text

Data-driven approach to designing compound words for continuous speech recognition by Saon, G., Padmanabhan, M.

“…We present a new approach to deriving compound words from a training corpus. The motivation for making compound words is because under some assumptions, speech…”
Get full text

Automatic speech recognition performance on a voicemail transcription task by Padmanabhan, M., Saon, G., Huang, J., Kingsbury, B., Mangu, L.

“…We report on the performance of automatic speech recognition (ASR) systems on voicemail transcription. Voicemail is spontaneous telephone speech recorded over…”
Get full text

Boosted MMI for model and feature-space discriminative training by Povey, D., Kanevsky, D., Kingsbury, B., Ramabhadran, B., Saon, G., Visweswariah, K.

“…We present a modified form of the maximum mutual information (MMI) objective function which gives improved results for discriminative training. The…”
Get full text

Bayesian Sensing Hidden Markov Models by Saon, G., Jen-Tzung Chien

“…In this paper, we introduce Bayesian sensing hidden Markov models (BS-HMMs) to represent sequential data based on a set of state-dependent basis vectors. The…”
Get full text

fMPE: discriminatively trained features for speech recognition by Povey, D., Kingsbury, B., Mangu, L., Saon, G., Soltau, H., Zweig, G.

“…MPE (minimum phone error) is a previously introduced technique for discriminative training of HMM parameters. fMPE applies the same objective function to the…”
Get full text

A Non-Linear Speaker Adaptation Technique using Kernel Ridge Regression by Saon, G.

“…We propose a non-linear model space transformation for speaker or environment adaptation based on weighted kernel ridge regression (KRR). The transformation is…”
Get full text

Eliminating inter-speaker variability prior to discriminant transforms by Saon, G., Padmanabhan, M., Gopinath, R.

“…This paper shows the impact of speaker normalization techniques, such as vocal tract length normalization (VTLN) and speaker-adaptive training (SAT), prior to…”
Get full text

The IBM 2004 conversational telephony system for rich transcription by Soltau, H., Kingsbury, B., Mangu, L., Povey, D., Saon, G., Zweig, G.

“…This paper describes the technical advances in IBM's conversational telephony submission to the DARPA-sponsored 2004 rich transcription evaluation (RT-04)…”
Get full text

The IBM Attila speech recognition toolkit by Soltau, H, Saon, G, Kingsbury, B

“…We describe the design of IBM's Attila speech recognition toolkit. We show how the combination of a highly modular and efficient library of low-level C++…”
Get full text

Off-line handwritten word recognition using a mixed HMM-MRF approach by Saon, G., Belaid, A.

“…Presents a 2D stochastic method for the recognition of unconstrained handwritten words in a small lexicon. The method is based on an efficient combination of…”
Get full text

The IBM 2006 Gale Arabic ASR System by Soltau, H., Saon, G., Kingsbury, B., Kuo, J., Mangu, L., Povey, D., Zweig, G.

“…This paper describes the advances made in IBM's Arabic broadcast news transcription system which was fielded in the 2006 GALE ASR and machine translation…”
Get full text

Large margin semi-tied covariance transforms for discriminative training by Saon, G., Povey, D., Soltau, H.

“…We discuss the applicability of large margin techniques to the problem of estimating linear transforms for discriminative training of a semi-tied covariance…”
Get full text

Linear feature space projections for speaker adaptation by Saon, G., Zweig, G., Padmanabhan, M.

“…We extend the well-known technique of constrained maximum likelihood linear regression (MLLR) to compute a projection (instead of a full rank transformation)…”
Get full text

Arc minimization in finite-state decoding graphs with cross-word acoustic context by Yvon, François, Zweig, Geoffrey, Saon, George

“…Recent approaches to large vocabulary decoding with weighted finite-state transducers have focused on the use of determinization and minimization algorithms to…”
Get full text

Digit recognition in noisy environments via a sequential GMM/SVM system by Fine, Shai, Saon, George, Gopinath, Ramesh A.

“…This paper exploits the fact that when GMM and SVM classifiers with roughly the same level of performance exhibit uncorrelated errors they can be combined to…”
Get full text

Robust speech recognition in Noisy Environments: The 2001 IBM spine evaluation system by Kingsbury, Brian, Saon, George, Mangu, Lidia, Padmanabhan, Mukund, Sarikaya, Ruhi

“…We report on the system IBM fielded in the second SPeech In Noisy Environments (SPINE-2) evaluation, conducted by the Naval Research Laboratory in October…”
Get full text

Dynamic network decoding revisited by Soltau, H., Saon, G.

“…We present a dynamic network decoder capable of using large cross-word context models and large n-gram histories. Our method for constructing the search…”
Get full text