System combination and score normalization for spoken term detection

Mamou, Jonathan; Cui, Jia; Cui, Xiaodong; Gales, Mark J. F.; Kingsbury, Brian; Knill, Kate; Mangu, Lidia; Nolden, David; Picheny, Michael; Ramabhadran, Bhuvana; Schlüter, Ralf; Sethy, Abhinav; Woodland, Philip C.

doi:10.1109/icassp.2013.6639278

Cited by 49 publications

(50 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The specific method used in this work, which is described in more detail below, is to (1) apply sum-toone normalization to each postings list, (2) combine the results using MTWV-weighted CombMNZ fusion, and finally (3) apply sum-toone normalization to the fused postings list to produce the final output. Experiments that led to this particular strategy are described in detail in [24].…”

Section: Score Normalization and System Combination For Abstractearchmentioning

confidence: 99%

“…Sum-to-one normalization does this, using the sum of the query detection scores as a proxy for query frequency. System combination is performed using an MTWV-weighted version [24] of the CombMNZ method [27]. First, detection scores in each postings list are weighted in proportion to the list's MTWV score on the tuning portion of the development set.…”

Section: Score Normalization and System Combination For Abstractearchmentioning

confidence: 99%

See 1 more Smart Citation

A high-performance Cantonese keyword search system

Kingsbury

Cui

et al. 2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

We present a system for keyword search on Cantonese conversational telephony audio, collected for the IARPA Babel program, that achieves good performance by combining postings lists produced by diverse speech recognition systems from three different research groups. We describe the keyword search task, the data on which the work was done, four different speech recognition systems, and our approach to system combination for keyword search. We show that the combination of four systems outperforms the best single system by 7%, achieving an actual term-weighted value of 0.517.

show abstract

Section: Score Normalization and System Combination For Abstractearchmentioning

confidence: 99%

Section: Score Normalization and System Combination For Abstractearchmentioning

confidence: 99%

A high-performance Cantonese keyword search system

Kingsbury

Cui

et al. 2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

show abstract

“…By fusing KWS results from diverse systems, we can usually get a much better KWS result. For fusing results of different systems, arithmetic-based fusion methods such as CombSum [1,2], CombMNZ [1,2], CombGMNZ [1], WCombMNZ [2] have been proved to be quite effective. Pham et al [3] proposed the system and keyword dependent fusion method SKDWCombMNZ in 2014, which ourperformed other arithmetic-based methods.…”

Section: Introductionmentioning

confidence: 99%

“…For the two measures, score normalization [2,7,8] has been proved to be essential. Keyword specific threshold (KST) normalization [9] and sum-to-one (STO) normalization [2] are the two mainstream score normalization methods. In our work, we compare the performance of the two methods when they are applied both before and after system fusion.…”

Section: Introductionmentioning

confidence: 99%

Improved system fusion for keyword search

Cai

Lü

et al. 2015

2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU)

View full text Add to dashboard Cite

It has been demonstrated that system fusion can significantly improve the performance of keyword search. In this paper, we compare the performance of several widely-used arithmeticbased fusion methods using different normalization pipeline and try to find the best pipeline. A novel arithmetic-based fusion method is proposed in this work. The method supplies a more effective way to incorporate the number of systems which have non-zero scores for a detection. When tested on the development test dataset of the OpenKWS15 Evaluation, the proposed method achieves the highest maximum termweighted value (MTWV) and actual term-weighted value (ATWV) among all other arithmetic-based fusion methods. Usually, discriminative fusion methods employing classifiers can outperform arithmetic-based fusion methods. A DNNbased fusion method is explored in this work. After wordburst information is added, the DNN-based fusion method outperforms all other methods. In addition, it is notable that our arithmetic-based method achieves the same MTWV as the DNN-based method.

show abstract

“…For target languages that are rare or spoken only in regions where collection is difficult or even impossible, LRs can be scarce or non-existent, imposing major, often irresolvable, constraints on training acoustic models in these languages. In dealing with under-resourced or zero-resource languages, research is focused on finding robust KWS solutions using techniques such as subspace-GMM acoustic modeling, [25] multiple system combination and score normalization, [26] and bootstrapping techniques utilizing multilanguage acoustic models and neural networks. [9,27,28] When adapting a PS KWS system to process data in an under-resourced language, it is possible to bypass the long and costly training process by utilizing phoneme acoustic models from accessible and well-trained languages.…”

Section: Abstractpotting In Under-resourced Languagesmentioning

confidence: 99%

Cross-language phoneme mapping for phonetic search keyword spotting in continuous speech of under-resourced languages

Tetariy

Bar-Yosef

Silber-Varod

et al. 2015

AIR

View full text Add to dashboard Cite

As automatic speech recognition-based applications become increasingly common in a wide variety of market segments, there is a growing need to support more languages. However, for many languages, the language resources needed to train speech recognition engines are either limited or completely non-existent, and the process of acquiring or constructing new language resources is both long and costly. This paper suggests a methodology that enables Phonetic Search Keyword Spotting to be implemented in a large speech database of any given under-resourced language using cross-language phoneme mappings to another language. The phoneme mapping enables a speech recognition engine from a sufficiently resourced and well-trained source language to be used for phoneme recognition in the new target language. The keyword search is then performed over a lattice of target language phonemes. Three cross-language phoneme mapping techniques are examined: knowledge-based, data-driven and phoneme recognition performance-based. The results suggest that Phonetic Search Keyword Spotting based on the cross-language phoneme mapping approach proposed herein can serve as a quick initial solution for validating keyword spotting applications in new, under-resourced languages.

show abstract

System combination and score normalization for spoken term detection

Cited by 49 publications

References 27 publications

A high-performance Cantonese keyword search system

A high-performance Cantonese keyword search system

Improved system fusion for keyword search

Cross-language phoneme mapping for phonetic search keyword spotting in continuous speech of under-resourced languages

Contact Info

Product

Resources

About