TY - GEN
T1 - Zero-shot audio classification with factored linear and nonlinear acoustic-semantic projections
AU - Xie, Huang
AU - Räsänen, Okko
AU - Virtanen, Tuomas
N1 - Funding Information:
The research leading to these results has received funding from the European Research Council under the European Unions H2020 Framework Programme through ERC Grant Agreement 637422 EVERYSOUND. OR was funded by Academy of Finland grant no. 314602.
Publisher Copyright:
© 2021 IEEE
JUFOID=57409
PY - 2021
Y1 - 2021
N2 - In this paper, we study zero-shot learning in audio classification through factored linear and nonlinear acoustic-semantic projections between audio instances and sound classes. Zero-shot learning in audio classification refers to classification problems that aim at recognizing audio instances of sound classes, which have no available training data but only semantic side information. In this paper, we address zero-shot learning by employing factored linear and nonlinear acoustic-semantic projections. We develop factored linear projections by applying rank decomposition to a bilinear model, and use nonlinear activation functions, such as tanh, to model the non-linearity between acoustic embeddings and semantic embeddings. Compared with the prior bilinear model, experimental results show that the proposed projection methods are effective for improving classification performance of zero-shot learning in audio classification.
AB - In this paper, we study zero-shot learning in audio classification through factored linear and nonlinear acoustic-semantic projections between audio instances and sound classes. Zero-shot learning in audio classification refers to classification problems that aim at recognizing audio instances of sound classes, which have no available training data but only semantic side information. In this paper, we address zero-shot learning by employing factored linear and nonlinear acoustic-semantic projections. We develop factored linear projections by applying rank decomposition to a bilinear model, and use nonlinear activation functions, such as tanh, to model the non-linearity between acoustic embeddings and semantic embeddings. Compared with the prior bilinear model, experimental results show that the proposed projection methods are effective for improving classification performance of zero-shot learning in audio classification.
KW - Acoustic-semantic projection
KW - Audio classification
KW - Zero-shot learning
U2 - 10.1109/ICASSP39728.2021.9414994
DO - 10.1109/ICASSP39728.2021.9414994
M3 - Conference contribution
AN - SCOPUS:85115058220
SN - 978-1-7281-7605-5
VL - 2021-June
T3 - Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing
SP - 326
EP - 330
BT - 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing
PB - IEEE
T2 - IEEE International Conference on Acoustics, Speech and Signal Processing
Y2 - 1 January 1900 through 1 January 2000
ER -