Keyword Spotting Using Support Vector Machines

Ayed, Yassine Ben; Fohr, Dominique; Haton, Jean Paul; Chollet, Gérard

doi:10.1007/3-540-46154-x_39

Cited by 17 publications

(10 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with the studies reported by others, such as Ou et al (2001); Ayed et al (2002);Wang et al (2012), the novel features in this work are those involving: Levenshtein distances, position, prosodic features, all the lexical features except the number of phones, and all the duration features except the duration of the detection.…”

Section: Features In Analysismentioning

confidence: 71%

“…These factors also influence the reliability of detections and hence can be utilized in estimating the confidence of detections. Research has been conducted on confidence estimation utilizing various informative factors, in both automatic speech recognition and keyword spotting, e.g., Rohlicek et al (1989); Cox and Rose (1996); Bergen and Ward (1997); Kemp and Schaaf (1997); Ou et al (2001); Ayed et al (2002); Jiang (2005), and various methods have been employed to combine the heterogeneous informative factors, including decision trees (DT), general linear models (GLMs), generalized additive models (GAMs) and multi-layer perceptrons (MLPs) (Chase, 1997;Gillick et al, 1997;Zhang and Rudnicky, 2001). It has been found that features derived from multiple sources -with appropriate normalization -can be combined to serve as a good measure of confidence, which can in turn be used to evaluate the correctness of a recognition hypothesis or a keyword detection.…”

Section: Motivation and Organization Of This Papermentioning

confidence: 99%

“…Ou et al (2001) employed word posterior likelihoods derived from keyword, anti-keyword and non-keyword models and duration as features in a neural network classifier for utterance verification within a filler model-based keyword spotting system whereas Ayed et al (2002) used information based on the number of frames, the phone posterior probability, the frame-based phone posterior probability and the duration-based phone posterior probability along with the number of phones as lexical feature in an SVM classifier for utterance verification in an LVCSR-based keyword spotting system.…”

Section: Feature Collectionmentioning

confidence: 99%

“…Among various discriminative models, MLPs (Mathan and Miclet, 1991;Weintraub et al, 1997;Ou et al, 2001;Vergyri et al, 2006), linear discriminative functions (Sukkar and Wilpon, 1993;Gillick et al, 1997;Kamppari and Hazen, 2000), GLMs and GAMs (Siu et al, 1997), DTs (Neti et al, 1997;Hauptmann et al, 1998) and SVMs (Zhang and Rudnicky, 2001;Ayed et al, 2002;Sudoh et al, 2006;Shafran et al, 2006) are the most popular. Comparisons between DTs, GLMs, GAMs and neural networks have been carried out by Chase (1997), and linear classifiers and neural networks are compared in .…”

Section: Feature Combinationmentioning

confidence: 99%

See 3 more Smart Citations

Feature analysis for discriminative confidence estimation in spoken term detection

Tejedor

Toledano

Wang

et al. 2014

Computer Speech & Language

View full text Add to dashboard Cite

El acceso a la versión del editor puede requerir la suscripción del recurso Access to the published version may require subscription AbstractDiscriminative confidence based on multi-layer perceptrons (MLPs) and multiple features has shown significant advantage compared to the widely used lattice-based confidence in spoken term detection (STD). Although the MLP-based framework can handle any features derived from a multitude of sources, choosing all possible features may lead to over complex models and hence less generality. In this paper, we design an extensive set of features and analyze their contribution to STD individually and as a group. The main goal is to choose a small set of features that are sufficiently informative while keeping the model simple and generalizable. We employ two established models to conduct the analysis: one is linear regression which targets for the most relevant features and the other is logistic linear regression which targets for the most discriminative features. We find the most informative features are comprised of those derived from diverse sources (ASR decoding, duration and lexical properties) and the two models deliver highly consistent feature ranks. STD experiments on both English and Spanish data demonstrate significant performance gains with the proposed feature sets.

show abstract

Section: Features In Analysismentioning

confidence: 71%

Section: Motivation and Organization Of This Papermentioning

confidence: 99%

Section: Feature Collectionmentioning

confidence: 99%

Section: Feature Combinationmentioning

confidence: 99%

See 2 more Smart Citations

Feature analysis for discriminative confidence estimation in spoken term detection

Tejedor

Toledano

Wang

et al. 2014

Computer Speech & Language

View full text Add to dashboard Cite

show abstract

“…This is most commonly done with Hidden Markov Models (HMM) [1,2]. However, the use of HMMs has various drawbacks, such as the need for an adequate "garbage model" to handle non-keyword speech.…”

Section: Introductionmentioning

confidence: 99%

Robust discriminative keyword spotting for emotionally colored spontaneous speech using bidirectional LSTM networks

Wöllmer

Eyben

Keshet

et al. 2009

2009 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

In this paper we propose a new technique for robust keyword spotting that uses bidirectional Long Short-Term Memory (BLSTM) recurrent neural nets to incorporate contextual information in speech decoding. Our approach overcomes the drawbacks of generative HMM modeling by applying a discriminative learning procedure that non-linearly maps speech features into an abstract vector space. By incorporating the outputs of a BLSTM network into the speech features, it is able to make use of past and future context for phoneme predictions. The robustness of the approach is evaluated on a keyword spotting task using the HUMAINE Sensitive Artificial Listener (SAL) database, which contains accented, spontaneous, and emotionally colored speech. The test is particularly stringent because the system is not trained on the SAL database, but only on the TIMIT corpus of read speech. We show that our method prevails over a discriminative keyword spotter without BLSTM-enhanced feature functions, which in turn has been proven to outperform HMM-based techniques.

show abstract

Cross-Context Linking Concepts Discovery in E-Government Literature

Cestnik

Kern²

2016

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Part 1: E-Government FoundationsInternational audienceTo conduct their business, organizations are nowadays challenged to handle huge amount of information from heterogeneous sources. Novel technologies can help them dealing with this delicate assignment. In this paper we describe an approach to document clustering and outlier detection that is regularly used to organize and summarize knowledge stored in huge amounts of documents in a government organization. The motivation for our preliminary study has been three-fold: first, to obtain an overview of the topics addressed in the recently published e-government papers, with the emphasis on identifying the shift of focus through the years; second, to form a collection of papers related to a preselected terms of interest in order to explore the characteristic keywords that discriminate this collection with respect to the rest of the documents; and third, to compare the papers that address a similar topic from two document sources and to show characteristic similarities and differences between the two origins, with a particular aim to identify outlier papers in each document source that are potentially worth for further exploration. As a document source for our study we used E-Government Reference Library of articles and PubMed. The presented case study results suggest that the document exploration supported by a document clustering tool can be more focused, efficient and effective

show abstract

Keyword Spotting Using Support Vector Machines

Cited by 17 publications

References 2 publications

Feature analysis for discriminative confidence estimation in spoken term detection

Feature analysis for discriminative confidence estimation in spoken term detection

Robust discriminative keyword spotting for emotionally colored spontaneous speech using bidirectional LSTM networks

Cross-Context Linking Concepts Discovery in E-Government Literature

Contact Info

Product

Resources

About