ACOUSTIC SCENE CLASSIFICATION USING AUTOENCODER

@inproceedings{Chen2017ACOUSTICSC,
  title={ACOUSTIC SCENE CLASSIFICATION USING AUTOENCODER},
  author={Xiaoou Chen and Deshun Yang},
  year={2017}
}
This report describes our contribution to the Acoustic Scene Classification (ASC) task of the 2017 IEEE AASP DCASE challenge[1]. We apply an Autoencoder to capture the discriminative information underlying the audio. Then, a Logistic Regression model is employed to recognize different scenes under the compressed representation. In order to boost the performance, we train models based on different channels from the original recordings and simply apply majority voting method on the predictions… CONTINUE READING

References

Publications referenced by this paper.
SHOWING 1-9 OF 9 REFERENCES

: a hybrid approach using binaural i - vectors and deep convolutional neural networks

  • M. Dorfer, G. Widmer
  • 2016

Similar Papers

Loading similar papers…