Abstract
This paper proposes a novel deep convolutional neural network (CNN), called sparse coding convolutional neural network (SC-CNN), to address the problem of sound event recognition and retrieval task. Unlike the general framework of a CNN, in which feature learning process is performed hierarchically, the proposed framework models the whole memorizing procedures in the human brain, including encoding, storage, and recollection. Sound data from the RWCP sound scene dataset with added noise from NOISEX-92 noise dataset are used to compare the performance of the proposed system with the state-of-the-art baselines. The experimental results indicated that the proposed SC-CNN outperformed the state-of-the-art systems in sound event recognition and retrieval. In the sound event recognition task, the proposed system achieved an accuracy of 94.6%, 100% and 100% under 0db, 10db and clean noise conditions, respectively. In the retrieval task, the proposed system improves the mAP rate of the general CNN by approximately 6%.
| Original language | English |
|---|---|
| Title of host publication | 2017 IEEE International Conference on Multimedia and Expo, ICME 2017 |
| Publisher | IEEE Computer Society |
| Pages | 589-594 |
| Number of pages | 6 |
| ISBN (Electronic) | 9781509060672 |
| DOIs | |
| Publication status | Published - 2017 Aug 28 |
| Event | 2017 IEEE International Conference on Multimedia and Expo, ICME 2017 - Hong Kong, Hong Kong Duration: 2017 Jul 10 → 2017 Jul 14 |
Publication series
| Name | Proceedings - IEEE International Conference on Multimedia and Expo |
|---|---|
| ISSN (Print) | 1945-7871 |
| ISSN (Electronic) | 1945-788X |
Other
| Other | 2017 IEEE International Conference on Multimedia and Expo, ICME 2017 |
|---|---|
| Country | Hong Kong |
| City | Hong Kong |
| Period | 17-07-10 → 17-07-14 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications
- Computer Science Applications
Cite this
}
Recognition and retrieval of sound events using sparse coding convolutional neural network. / Wang, Chien Yao; Santoso, Andri; Mathulaprangsan, Seksan; Chiang, Chin Chin; Wu, Chung-Hsien; Wang, Jia Ching.
2017 IEEE International Conference on Multimedia and Expo, ICME 2017. IEEE Computer Society, 2017. p. 589-594 8019552 (Proceedings - IEEE International Conference on Multimedia and Expo).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Recognition and retrieval of sound events using sparse coding convolutional neural network
AU - Wang, Chien Yao
AU - Santoso, Andri
AU - Mathulaprangsan, Seksan
AU - Chiang, Chin Chin
AU - Wu, Chung-Hsien
AU - Wang, Jia Ching
PY - 2017/8/28
Y1 - 2017/8/28
N2 - This paper proposes a novel deep convolutional neural network (CNN), called sparse coding convolutional neural network (SC-CNN), to address the problem of sound event recognition and retrieval task. Unlike the general framework of a CNN, in which feature learning process is performed hierarchically, the proposed framework models the whole memorizing procedures in the human brain, including encoding, storage, and recollection. Sound data from the RWCP sound scene dataset with added noise from NOISEX-92 noise dataset are used to compare the performance of the proposed system with the state-of-the-art baselines. The experimental results indicated that the proposed SC-CNN outperformed the state-of-the-art systems in sound event recognition and retrieval. In the sound event recognition task, the proposed system achieved an accuracy of 94.6%, 100% and 100% under 0db, 10db and clean noise conditions, respectively. In the retrieval task, the proposed system improves the mAP rate of the general CNN by approximately 6%.
AB - This paper proposes a novel deep convolutional neural network (CNN), called sparse coding convolutional neural network (SC-CNN), to address the problem of sound event recognition and retrieval task. Unlike the general framework of a CNN, in which feature learning process is performed hierarchically, the proposed framework models the whole memorizing procedures in the human brain, including encoding, storage, and recollection. Sound data from the RWCP sound scene dataset with added noise from NOISEX-92 noise dataset are used to compare the performance of the proposed system with the state-of-the-art baselines. The experimental results indicated that the proposed SC-CNN outperformed the state-of-the-art systems in sound event recognition and retrieval. In the sound event recognition task, the proposed system achieved an accuracy of 94.6%, 100% and 100% under 0db, 10db and clean noise conditions, respectively. In the retrieval task, the proposed system improves the mAP rate of the general CNN by approximately 6%.
UR - http://www.scopus.com/inward/record.url?scp=85030237057&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85030237057&partnerID=8YFLogxK
U2 - 10.1109/ICME.2017.8019552
DO - 10.1109/ICME.2017.8019552
M3 - Conference contribution
T3 - Proceedings - IEEE International Conference on Multimedia and Expo
SP - 589
EP - 594
BT - 2017 IEEE International Conference on Multimedia and Expo, ICME 2017
PB - IEEE Computer Society
ER -