Trajectory Clustering for Solving the Trajectory Folding Problem in Automatic Speech Recognition

Han, Yan; Veth, Johan de; Boves, L.W.J.

doi:10.1109/tasl.2007.894529

Cited by 12 publications

(14 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our spectral reduction measure is susceptible to the well-known trajectory folding problem (Han et al, 2007); different tokens taking different trajectories through the acoustic space may end up with identical log-likelihoods, even if their trajectories make very different auditory impres sions. This is yet another reason why it may not be appro priate to map multidimensional acoustic reduction to a real number.…”

Section: Discussionmentioning

confidence: 99%

Analysis of acoustic reduction using spectral similarity measures

Hämäläinen

Gubian

Bosch

et al. 2009

The Journal of the Acoustical Society of America

Self Cite

View full text Add to dashboard Cite

Articulatory and acoustic reduction can manifest itself in the temporal and spectral domains. This study introduces a measure of spectral reduction, which is based on the speech decoding techniques commonly used in automatic speech recognizers. Using data for four frequent Dutch affixes from a large corpus of spontaneous face-to-face conversations, it builds on an earlier study examining the effects of lexical frequency on durational reduction in spoken Dutch [Pluymaekers, M. et al. (2005). J. Acoust. Soc. Am. 118, [2561][2562][2563][2564][2565][2566][2567][2568][2569], and compares the proposed measure of spectral reduction with duration as a measure of reduction. The results suggest that the spectral reduction scores capture other aspects of reduction than duration. While duration can-albeit to a moderate degreebe predicted by a number of linguistically motivated variables (such as word frequency, segmental context, and speech rate), the spectral reduction scores cannot. This suggests that the spectral reduction scores capture information that is not directly accounted for by the linguistically motivated variables. The results also show that the spectral reduction scores are able to predict a substantial amount of the variation in duration that the linguistically motivated variables do not account for.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of acoustic reduction using spectral similarity measures

Hämäläinen

Gubian

Bosch

et al. 2009

The Journal of the Acoustical Society of America

Self Cite

View full text Add to dashboard Cite

show abstract

“…The method of Han et al [7] does trajectory clustering by using the mixture model for automatic speech recognition. They report on the general problem of different initial cluster assignments leading to different EM clusters.…”

Section: Related Workmentioning

confidence: 99%

“…For each subsequent frame these parameters are predicted with the help of Kalman filter. Unlike the method of Han et al [7], which utilises splitting during initialisation, Xiong et al integrate "split", "merge" and "delete" operation into their dynamic Gaussian mixture model.…”

Section: Related Workmentioning

confidence: 99%

“…Second, a model over all data may smooth away concept drifts that are valuable to the observer. Some later works, like [7,14] address the first problem by adapting the model as time progresses and new data arrive. However, the problem of adapting to drift is treated in a rather rudimentary way, namely by considering only special aspects of drift (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Online Clustering of High-Dimensional Trajectories under Concept Drift

Krempl

Siddiqui

Spiliopoulou

2011

Machine Learning and Knowledge Discovery in Databases

View full text Add to dashboard Cite

Abstract. Historical transaction data are collected in many applications, e.g., patient histories recorded by physicians and customer transactions collected by companies. An important question is the learning of models upon the primary objects (patients, customers) rather than the transactions, especially when these models are subjected to drift.We address this problem by combining advances of online clustering on multivariate data with the trajectory mining paradigm. We model the measurements of each individual primary object (e.g. its transactions), taken at irregular time intervals, as a trajectory in a high-dimensional feature space. Then, we cluster individuals with similar trajectories to identify sub-populations that evolve similarly, e.g. groups of customers that evolve similarly or groups of employees that have similar careers.We assume that the multivariate trajectories are generated by drifting Gaussian Mixture Models. We study (i) an EM-based approach that clusters these trajectories incrementally as a reference method that has access to all the data for learning, and propose (ii) an online algorithm based on a Kalman filter that efficiently tracks the trajectories of Gaussian clusters. We show that while both methods approximate the reference well, the algorithm based on a Kalman filter is faster by one order of magnitude compared to the EM-based approach.

show abstract

“…Therefore, we must conclude that multi-path syllable models, however they may be initialised and trained, may not be the way towards solving the pronunciation variation problem in ASR. Using the acoustic variation in speech as the basis for constructing parametric models of speech (Deng et al, 2006;Han et al, 2007;Zen et al, 2007) will not solve the context modelling problem either. It may well aggravate the problem because it is difficult to link bottom-up acoustic variation to the lexicon and the language model.…”

Section: Directions For Future Researchmentioning

confidence: 99%