Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | 2019 AES International Conference on Audio Forensics |
| Herausgeber/-innen | Durand Begault, Douglas Lacey |
| Herausgeber (Verlag) | Audio Engineering Society |
| ISBN (elektronisch) | 9781942220282 |
| Publikationsstatus | Veröffentlicht - 8 Juni 2019 |
| Veranstaltung | 2019 AES International Conference on Audio Forensics - Porto, Portugal Dauer: 18 Juni 2019 → 20 Juni 2019 |
Publikationsreihe
| Name | Proceedings of the AES International Conference |
|---|---|
| Band | 2019-June |
Abstract
Gunshot classification in audio files is used in forensics, surveillance, and multimedia analysis. In this contribution we show that it is possible to use data augmentation in order to enlarge the training set of a rare event like a gunshot with artificial data based on a simple but sufficient model, and a database of room impulse responses. The results indicate that the enlarged database increases the accuracy in a classification task significantly, even if no real data is used for training at all.
ASJC Scopus Sachgebiete
- Ingenieurwesen (insg.)
- Elektrotechnik und Elektronik
- Physik und Astronomie (insg.)
- Akustik und Ultraschall
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
2019 AES International Conference on Audio Forensics. Hrsg. / Durand Begault; Douglas Lacey. Audio Engineering Society, 2019. (Proceedings of the AES International Conference; Band 2019-June).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Improved gunshot classification by using artificial data
AU - Busse, Christian
AU - Krause, Thomas
AU - Ostermann, Jörn
AU - Bitzer, Jörg
PY - 2019/6/8
Y1 - 2019/6/8
N2 - Gunshot classification in audio files is used in forensics, surveillance, and multimedia analysis. In this contribution we show that it is possible to use data augmentation in order to enlarge the training set of a rare event like a gunshot with artificial data based on a simple but sufficient model, and a database of room impulse responses. The results indicate that the enlarged database increases the accuracy in a classification task significantly, even if no real data is used for training at all.
AB - Gunshot classification in audio files is used in forensics, surveillance, and multimedia analysis. In this contribution we show that it is possible to use data augmentation in order to enlarge the training set of a rare event like a gunshot with artificial data based on a simple but sufficient model, and a database of room impulse responses. The results indicate that the enlarged database increases the accuracy in a classification task significantly, even if no real data is used for training at all.
UR - http://www.scopus.com/inward/record.url?scp=85071199734&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85071199734
T3 - Proceedings of the AES International Conference
BT - 2019 AES International Conference on Audio Forensics
A2 - Begault, Durand
A2 - Lacey, Douglas
PB - Audio Engineering Society
T2 - 2019 AES International Conference on Audio Forensics
Y2 - 18 June 2019 through 20 June 2019
ER -