Topic dependent language modelling for spoken term detection. Kalantari, S., Dean, D., Sridharan, S., & Wallace, R. In 2014 22nd European Signal Processing Conference (EUSIPCO), pages 949-953, Sep., 2014.
Paper abstract bibtex This paper investigates the effect of topic dependent language models (TDLM) on phonetic spoken term detection (STD) using dynamic match lattice spotting (DMLS). Phonetic STD consists of two steps: indexing and search. The accuracy of indexing audio segments into phone sequences using phone recognition methods directly affects the accuracy of the final STD system. If the topic of a document in known, recognizing the spoken words and indexing them to an intermediate representation is an easier task and consequently, detecting a search word in it will be more accurate and robust. In this paper, we propose the use of TDLMs in the indexing stage to improve the accuracy of STD in situations where the topic of the audio document is known in advance. It is shown that using TDLMs instead of the traditional general language model (GLM) improves STD performance according to figure of merit (FOM) criteria.
@InProceedings{6952309,
author = {S. Kalantari and D. Dean and S. Sridharan and R. Wallace},
booktitle = {2014 22nd European Signal Processing Conference (EUSIPCO)},
title = {Topic dependent language modelling for spoken term detection},
year = {2014},
pages = {949-953},
abstract = {This paper investigates the effect of topic dependent language models (TDLM) on phonetic spoken term detection (STD) using dynamic match lattice spotting (DMLS). Phonetic STD consists of two steps: indexing and search. The accuracy of indexing audio segments into phone sequences using phone recognition methods directly affects the accuracy of the final STD system. If the topic of a document in known, recognizing the spoken words and indexing them to an intermediate representation is an easier task and consequently, detecting a search word in it will be more accurate and robust. In this paper, we propose the use of TDLMs in the indexing stage to improve the accuracy of STD in situations where the topic of the audio document is known in advance. It is shown that using TDLMs instead of the traditional general language model (GLM) improves STD performance according to figure of merit (FOM) criteria.},
keywords = {indexing;speech recognition;topic dependent language modelling;spoken term detection;TDLM;STD;dynamic match lattice spotting;phone sequences;phone recognition methods;general language model;GLM;figure of merit criteria;indexing stage;search stage;Indexing;Accuracy;Speech;Lattices;Speech recognition;Hidden Markov models;spoken term detection;language modelling;indexing},
issn = {2076-1465},
month = {Sep.},
url = {https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569926121.pdf},
}
Downloads: 0
{"_id":"qG3ro47m3PLHurqFt","bibbaseid":"kalantari-dean-sridharan-wallace-topicdependentlanguagemodellingforspokentermdetection-2014","authorIDs":[],"author_short":["Kalantari, S.","Dean, D.","Sridharan, S.","Wallace, R."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["S."],"propositions":[],"lastnames":["Kalantari"],"suffixes":[]},{"firstnames":["D."],"propositions":[],"lastnames":["Dean"],"suffixes":[]},{"firstnames":["S."],"propositions":[],"lastnames":["Sridharan"],"suffixes":[]},{"firstnames":["R."],"propositions":[],"lastnames":["Wallace"],"suffixes":[]}],"booktitle":"2014 22nd European Signal Processing Conference (EUSIPCO)","title":"Topic dependent language modelling for spoken term detection","year":"2014","pages":"949-953","abstract":"This paper investigates the effect of topic dependent language models (TDLM) on phonetic spoken term detection (STD) using dynamic match lattice spotting (DMLS). Phonetic STD consists of two steps: indexing and search. The accuracy of indexing audio segments into phone sequences using phone recognition methods directly affects the accuracy of the final STD system. If the topic of a document in known, recognizing the spoken words and indexing them to an intermediate representation is an easier task and consequently, detecting a search word in it will be more accurate and robust. In this paper, we propose the use of TDLMs in the indexing stage to improve the accuracy of STD in situations where the topic of the audio document is known in advance. It is shown that using TDLMs instead of the traditional general language model (GLM) improves STD performance according to figure of merit (FOM) criteria.","keywords":"indexing;speech recognition;topic dependent language modelling;spoken term detection;TDLM;STD;dynamic match lattice spotting;phone sequences;phone recognition methods;general language model;GLM;figure of merit criteria;indexing stage;search stage;Indexing;Accuracy;Speech;Lattices;Speech recognition;Hidden Markov models;spoken term detection;language modelling;indexing","issn":"2076-1465","month":"Sep.","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569926121.pdf","bibtex":"@InProceedings{6952309,\n author = {S. Kalantari and D. Dean and S. Sridharan and R. Wallace},\n booktitle = {2014 22nd European Signal Processing Conference (EUSIPCO)},\n title = {Topic dependent language modelling for spoken term detection},\n year = {2014},\n pages = {949-953},\n abstract = {This paper investigates the effect of topic dependent language models (TDLM) on phonetic spoken term detection (STD) using dynamic match lattice spotting (DMLS). Phonetic STD consists of two steps: indexing and search. The accuracy of indexing audio segments into phone sequences using phone recognition methods directly affects the accuracy of the final STD system. If the topic of a document in known, recognizing the spoken words and indexing them to an intermediate representation is an easier task and consequently, detecting a search word in it will be more accurate and robust. In this paper, we propose the use of TDLMs in the indexing stage to improve the accuracy of STD in situations where the topic of the audio document is known in advance. It is shown that using TDLMs instead of the traditional general language model (GLM) improves STD performance according to figure of merit (FOM) criteria.},\n keywords = {indexing;speech recognition;topic dependent language modelling;spoken term detection;TDLM;STD;dynamic match lattice spotting;phone sequences;phone recognition methods;general language model;GLM;figure of merit criteria;indexing stage;search stage;Indexing;Accuracy;Speech;Lattices;Speech recognition;Hidden Markov models;spoken term detection;language modelling;indexing},\n issn = {2076-1465},\n month = {Sep.},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569926121.pdf},\n}\n\n","author_short":["Kalantari, S.","Dean, D.","Sridharan, S.","Wallace, R."],"key":"6952309","id":"6952309","bibbaseid":"kalantari-dean-sridharan-wallace-topicdependentlanguagemodellingforspokentermdetection-2014","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2014/html/papers/1569926121.pdf"},"keyword":["indexing;speech recognition;topic dependent language modelling;spoken term detection;TDLM;STD;dynamic match lattice spotting;phone sequences;phone recognition methods;general language model;GLM;figure of merit criteria;indexing stage;search stage;Indexing;Accuracy;Speech;Lattices;Speech recognition;Hidden Markov models;spoken term detection;language modelling;indexing"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2014url.bib","creationDate":"2021-02-13T17:43:41.659Z","downloads":0,"keywords":["indexing;speech recognition;topic dependent language modelling;spoken term detection;tdlm;std;dynamic match lattice spotting;phone sequences;phone recognition methods;general language model;glm;figure of merit criteria;indexing stage;search stage;indexing;accuracy;speech;lattices;speech recognition;hidden markov models;spoken term detection;language modelling;indexing"],"search_terms":["topic","dependent","language","modelling","spoken","term","detection","kalantari","dean","sridharan","wallace"],"title":"Topic dependent language modelling for spoken term detection","year":2014,"dataSources":["A2ezyFL6GG6na7bbs","oZFG3eQZPXnykPgnE"]}