Speech Keyword Spotting with Rule Based Segmentation

Greibus, Mindaugas; Telksnys, Laimutis

doi:10.1007/978-3-642-41947-8_17

Cited by 3 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the experiment was created speech corpora as described above with the phrase: "What is the time?". Two segmentation algorithms were used [15]: Multi Feature Extremum and Rule Based Segmentation (MEaRBS) and the threshold algorithm. It will be shown that both algorithms showed similar results both with synthetic and natural speech, but their quantitative segmentation results allow identifying the better one.…”

Section: Segmentation Results Evaluation Andmentioning

confidence: 99%

Segmentation Analysis using Synthetic Speech Signals

Greibus

Telksnys

2012

ElAEE

View full text Add to dashboard Cite

There is a need of certain number of speech examples to solve real life tasks. A natural speech corpora may suffer from limited phoneme, word or phrase combinations. On the other hand an experimenter can create huge speech corpora using synthetic speech that reflect investigative speech cases that it is similar enough to natural speech. The usage of synthetic and natural speech corpora for speech segmentation algorithm comparison is presented in this paper. The adequateness of synthetic and natural corpora criteria for speech segmentation was proved. The experiments results showed that synthetic signals can be used for speech algorithm research.

show abstract

Section: Segmentation Results Evaluation Andmentioning

confidence: 99%

Segmentation Analysis using Synthetic Speech Signals

Greibus

Telksnys

2012

ElAEE

View full text Add to dashboard Cite

show abstract

“…Therefore, queries can be directly compared to the segmented word images [1]. Segmentation-Free Approaches, that do not have a prior information about the document layout and resemble template matching approaches, which try to find parts of the document that match a model/template of the query [2].…”

Section: Introductionmentioning

confidence: 99%

Keyword Spotting in Handwritten Documents Using Projections of Oriented Gradients

Retsinas

Louloudis

Stamatopoulos

et al. 2016

2016 12th IAPR Workshop on Document Analysis Systems (DAS)

View full text Add to dashboard Cite

In this paper, we present a novel approach for segmentation-based handwritten keyword spotting. The proposed approach relies upon the extraction of a simple yet efficient descriptor which is based on projections of oriented gradients. To this end, a global and a local word image descriptors, together with their combination, are proposed. Retrieval is performed using to the euclidean distance between the descriptors of a query image and the segmented word images. The proposed methods have been evaluated on the dataset of the ICFHR 2014 Competition on handwritten keyword spotting. Experimental results prove the efficiency of the proposed methods compared to several state-of-the-art techniques.

show abstract

Learning Efficient Representations for Keyword Spotting with Triplet Loss

Vygon

Mikhaylovskiy

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In the past few years, triplet loss-based metric embeddings have become a de-facto standard for several important computer vision problems, most notably, person reidentification. On the other hand, in the area of speech recognition the metric embeddings generated by the triplet loss are rarely used even for classification problems. We fill this gap showing that a combination of two representation learning techniques: a triplet loss-based embedding and a variant of kNN for classification instead of cross-entropy loss significantly (by 26% to 38%) improves the classification accuracy for convolutional networks on a LibriSpeech-derived LibriWords datasets. To do so, we propose a novel phonetic similarity based triplet mining approach. We also match the current best published SOTA for Google Speech Commands dataset V2 10+2-class classification with an architecture that is about 6 times more compact and improve the current best published SOTA for 35class classification on Google Speech Commands dataset V2 by over 40%. 1

show abstract

Speech Keyword Spotting with Rule Based Segmentation

Cited by 3 publications

References 16 publications

Segmentation Analysis using Synthetic Speech Signals

Segmentation Analysis using Synthetic Speech Signals

Keyword Spotting in Handwritten Documents Using Projections of Oriented Gradients

Learning Efficient Representations for Keyword Spotting with Triplet Loss

Contact Info

Product

Resources

About