Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions

Houssemeddine Khemiri; Gérard Chollet; Dijana Petrovska-Delacrétaz

doi:10.1109/CBMI.2011.5972549

Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions

Houssemeddine Khemiri
, Gérard Chollet
, Dijana Petrovska-Delacrétaz

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This paper describes an audio indexing system to search for known advertisements in radio broadcast streams, using automatically acquired segmental units. These segmental units called ALISP units are acquired automatically using temporal decomposition and vector quantization and modeled by Hidden Markov Models (HMMs). To detect commercials, ALISP transcriptions of reference advertisements are compared to those of radio stream using the Leven-shtein distance. The system is described and evaluated using broadcast streams provided by YACAST. On a set of 802 advertisements we achieve a mean precision of 95% with the corresponding recall value of 97%. The results show that the system is robust in situations where the advertisement to detect is stretched or suffer from time distortions. Moreover, this system allowed us to detect some annotation errors.

Original language	English
Title of host publication	CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing
Pages	223-228
Number of pages	6
DOIs	https://doi.org/10.1109/CBMI.2011.5972549
Publication status	Published - 6 Sept 2011
Event	9th International Workshop on Content-Based Multimedia Indexing, CBMi 2011 - Madrid, Spain Duration: 13 Jun 2011 → 15 Jun 2011

Publication series

Name	Proceedings - International Workshop on Content-Based Multimedia Indexing
ISSN (Print)	1949-3991

Conference

Conference	9th International Workshop on Content-Based Multimedia Indexing, CBMi 2011
Country/Territory	Spain
City	Madrid
Period	13/06/11 → 15/06/11

Access to Document

10.1109/CBMI.2011.5972549

Cite this

Khemiri, H., Chollet, G., & Petrovska-Delacrétaz, D. (2011). Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions. In CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing (pp. 223-228). Article 5972549 (Proceedings - International Workshop on Content-Based Multimedia Indexing). https://doi.org/10.1109/CBMI.2011.5972549

@inproceedings{35a8216c42b048daa02154ec955c736f,

title = "Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions",

abstract = "This paper describes an audio indexing system to search for known advertisements in radio broadcast streams, using automatically acquired segmental units. These segmental units called ALISP units are acquired automatically using temporal decomposition and vector quantization and modeled by Hidden Markov Models (HMMs). To detect commercials, ALISP transcriptions of reference advertisements are compared to those of radio stream using the Leven-shtein distance. The system is described and evaluated using broadcast streams provided by YACAST. On a set of 802 advertisements we achieve a mean precision of 95\% with the corresponding recall value of 97\%. The results show that the system is robust in situations where the advertisement to detect is stretched or suffer from time distortions. Moreover, this system allowed us to detect some annotation errors.",

author = "Houssemeddine Khemiri and G{\'e}rard Chollet and Dijana Petrovska-Delacr{\'e}taz",

year = "2011",

month = sep,

day = "6",

doi = "10.1109/CBMI.2011.5972549",

language = "English",

isbn = "9781612844336",

series = "Proceedings - International Workshop on Content-Based Multimedia Indexing",

pages = "223--228",

booktitle = "CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing",

note = "9th International Workshop on Content-Based Multimedia Indexing, CBMi 2011 ; Conference date: 13-06-2011 Through 15-06-2011",

}

Khemiri, H, Chollet, G & Petrovska-Delacrétaz, D 2011, Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions. in CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing., 5972549, Proceedings - International Workshop on Content-Based Multimedia Indexing, pp. 223-228, 9th International Workshop on Content-Based Multimedia Indexing, CBMi 2011, Madrid, Spain, 13/06/11. https://doi.org/10.1109/CBMI.2011.5972549

Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions. / Khemiri, Houssemeddine; Chollet, Gérard; Petrovska-Delacrétaz, Dijana.
CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing. 2011. p. 223-228 5972549 (Proceedings - International Workshop on Content-Based Multimedia Indexing).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions

AU - Khemiri, Houssemeddine

AU - Chollet, Gérard

AU - Petrovska-Delacrétaz, Dijana

PY - 2011/9/6

Y1 - 2011/9/6

N2 - This paper describes an audio indexing system to search for known advertisements in radio broadcast streams, using automatically acquired segmental units. These segmental units called ALISP units are acquired automatically using temporal decomposition and vector quantization and modeled by Hidden Markov Models (HMMs). To detect commercials, ALISP transcriptions of reference advertisements are compared to those of radio stream using the Leven-shtein distance. The system is described and evaluated using broadcast streams provided by YACAST. On a set of 802 advertisements we achieve a mean precision of 95% with the corresponding recall value of 97%. The results show that the system is robust in situations where the advertisement to detect is stretched or suffer from time distortions. Moreover, this system allowed us to detect some annotation errors.

AB - This paper describes an audio indexing system to search for known advertisements in radio broadcast streams, using automatically acquired segmental units. These segmental units called ALISP units are acquired automatically using temporal decomposition and vector quantization and modeled by Hidden Markov Models (HMMs). To detect commercials, ALISP transcriptions of reference advertisements are compared to those of radio stream using the Leven-shtein distance. The system is described and evaluated using broadcast streams provided by YACAST. On a set of 802 advertisements we achieve a mean precision of 95% with the corresponding recall value of 97%. The results show that the system is robust in situations where the advertisement to detect is stretched or suffer from time distortions. Moreover, this system allowed us to detect some annotation errors.

U2 - 10.1109/CBMI.2011.5972549

DO - 10.1109/CBMI.2011.5972549

M3 - Conference contribution

AN - SCOPUS:80052282402

SN - 9781612844336

T3 - Proceedings - International Workshop on Content-Based Multimedia Indexing

SP - 223

EP - 228

BT - CBMi 2011 - 9th International Workshop on Content-Based Multimedia Indexing

T2 - 9th International Workshop on Content-Based Multimedia Indexing, CBMi 2011

Y2 - 13 June 2011 through 15 June 2011

ER -

Automatic detection of known advertisements in radio broadcast with data-driven ALISP transcriptions

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this