Progress in dynamic programming search for LVCSR

Ney, Hermann; Ortmanns, S.

doi:10.1109/5.880081

Cited by 101 publications

(61 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The number of insertion, deletion and substitution errors is computed using the best alignment between two token sequences: the manually aligned (reference) and the recognized (test). An alignment resulting from search strategies based on dynamic programming is normally used successfully for a large number of speech recognition tasks, (Ney & Ortmanns, 2000). Speech recognition toolkits, such as HTK, (Young et al, 2006), include tools to compute accuracy and related measures on the basis of the transcribed data and recognition outputs using this dynamic programming algorithm.…”

Section: Standard Evaluation Phone Recognition Metricsmentioning

confidence: 99%

Phoneme Recognition on the TIMIT Database

Lopes¹,

Perdigão²

2011

Speech Technologies

View full text Add to dashboard Cite

Section: Standard Evaluation Phone Recognition Metricsmentioning

confidence: 99%

Phoneme Recognition on the TIMIT Database

Lopes¹,

Perdigão²

2011

Speech Technologies

View full text Add to dashboard Cite

“…That is, a standard HMM for one sub-interaction and a dynamic Bayesian network for another sub-interaction can be linked together in a network to represent interactions which include those sub-interactions. We also proposed a onepass DP search [36] based inference algorithm for NDPM. In the proposed NDPM, it is natural to build a cyclic model to represent repetitive sub-interactions in a complex interaction without increasing the computational cost and redesigning the interaction models.…”

Section: Conclusion and Future Researchmentioning

confidence: 99%

“…The solution is based on the best model and state sequence that jointly maximize the likelihood along with the segmental sequence. In order to tackle this problem, we exploit the one-pass DP search [36].…”

Section: B Inference In Ndpmmentioning

confidence: 99%

A Network of Dynamic Probabilistic Models for Human Interaction Analysis

Suk

Jain

Lee

2011

IEEE Trans. Circuits Syst. Video Technol.

View full text Add to dashboard Cite

Abstract-We propose a novel method of analyzing human interactions based on the walking trajectories of human subjects, which provide elementary and necessary components for understanding and interpretation of complex human interactions in visual surveillance tasks. Our principal assumption is that an interaction episode is composed of meaningful small unit interactions, which we call "sub-interactions." We model each sub-interaction by a dynamic probabilistic model and propose a modified factorial hidden Markov model (HMM) with factored observations. The complete interaction is represented with a network of dynamic probabilistic models (DPMs) by an ordered concatenation of sub-interaction models. The rationale for this approach is that it is more effective in utilizing common components, i.e., sub-interaction models, to describe complex interaction patterns. By assembling these sub-interaction models in a network, possibly with a mixture of different types of DPMs, such as standard HMMs, variants of HMMs, dynamic Bayesian networks, and so on, we can design a robust model for the analysis of human interactions. We show the feasibility and effectiveness of the proposed method by analyzing the structure of network of DPMs and its success on four different databases: a self-collected dataset, Tsinghua University's dataset, the public domain CAVIAR dataset, and the Edinburgh Informatics Forum Pedestrian dataset.Index Terms-Dynamic Bayesian network, human interaction analysis, network of dynamic probabilistic models, subinteractions, video surveillance.

show abstract

“…12: end if 13: erase the word graph and delete the dead paths [7] is. After the UWGE process, the lattice is turned into a sausage-like form and keeps every decoding hypothesis.…”

Section: B the Unconstrained Word Graph Expansion Algorithmmentioning

confidence: 99%

Unconstrained Word Graph Based Keyword Spotting

Zhang¹,

Si²,

Liu³

et al. 2013

Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013)

View full text Add to dashboard Cite

Abstract-The performance of keyword spotting system suffers severe degradation when the index stage is so fast that the lattice may lose lots of information to retrieve the spoken terms . In this paper , We focus on this problem and present an approach named unconstraint word graph expansion (UWGE) to keep the pruned hypotheses which are discarded in the decoding procedure but may contain correct hypotheses. The proposed approach is to eliminate the N-gram language model state limitation of lattice and reconstruct lattice to unconstrained word graph. On two Mandarin conversation telephone speech sets, we compare performance using UWGE with that on traditional trigram lattice , and our approach gives satisfying performance gains over trigram lattice. We also show the relationship between the performance and the system speed based on this approach.

show abstract

Progress in dynamic programming search for LVCSR

Cited by 101 publications

References 55 publications

Phoneme Recognition on the TIMIT Database

Phoneme Recognition on the TIMIT Database

A Network of Dynamic Probabilistic Models for Human Interaction Analysis

Unconstrained Word Graph Based Keyword Spotting

Contact Info

Product

Resources

About