TY - GEN
T1 - Audio source segmentation using spectral correlation features for automatic indexing of broadcast news
AU - Matsunaga, Shoichi
AU - Mizuno, Osamu
AU - Ohtsuki, Katsutoshi
AU - Hayashi, Yoshihiko
PY - 2015/4/3
Y1 - 2015/4/3
N2 - This paper proposes a new segmentation procedure to detect audio source intervals for automatic indexing of broadcast news. The procedure is composed of an audio source detection part and a part that smoothes the detected sequences. The detection part uses three new acoustic feature parameters that are based on spectral cross-correlation: spectral stability, white noise similarity, and sound spectral shape. These parameters make it possible to capture the audio sources more accurately than can be done with conventional parameters. The smoothing part has a new merging method that drops erroneous detection results of short duration. Audio source classification experiments are conducted on broadcast news segments. Performance is increased by 6.6% when the proposed parameters are used and by 3.1% when the proposed merging method is used, showing the usefulness of our approach. Experiments confirm the impact of this proposal on broadcast news indexing.
AB - This paper proposes a new segmentation procedure to detect audio source intervals for automatic indexing of broadcast news. The procedure is composed of an audio source detection part and a part that smoothes the detected sequences. The detection part uses three new acoustic feature parameters that are based on spectral cross-correlation: spectral stability, white noise similarity, and sound spectral shape. These parameters make it possible to capture the audio sources more accurately than can be done with conventional parameters. The smoothing part has a new merging method that drops erroneous detection results of short duration. Audio source classification experiments are conducted on broadcast news segments. Performance is increased by 6.6% when the proposed parameters are used and by 3.1% when the proposed merging method is used, showing the usefulness of our approach. Experiments confirm the impact of this proposal on broadcast news indexing.
UR - http://www.scopus.com/inward/record.url?scp=84979896262&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84979896262&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84979896262
T3 - European Signal Processing Conference
SP - 2103
EP - 2106
BT - 2004 12th European Signal Processing Conference, EUSIPCO 2004
PB - European Signal Processing Conference, EUSIPCO
T2 - 12th European Signal Processing Conference, EUSIPCO 2004
Y2 - 6 September 2004 through 10 September 2004
ER -