Autoencoder-based codebook for Bag-of-Audio-Words raises CCC for arousal from 0.225 to 0.322 and valence from 0.244 to 0.368 on AVEC 2017 audio data versus standard BoW.
A Unified Deep Neural Network for Speaker and Language Recognition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Learned feature representations and sub-phoneme posteriors from Deep Neural Networks (DNNs) have been used separately to produce significant performance gains for speaker and language recognition tasks. In this work we show how these gains are possible using a single DNN for both speaker and language recognition. The unified DNN approach is shown to yield substantial performance improvements on the the 2013 Domain Adaptation Challenge speaker recognition task (55% reduction in EER for the out-of-domain condition) and on the NIST 2011 Language Recognition Evaluation (48% reduction in EER for the 30s test condition).
fields
eess.AS 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Bag-of-Audio-Words based on Autoencoder Codebook for Continuous Emotion Prediction
Autoencoder-based codebook for Bag-of-Audio-Words raises CCC for arousal from 0.225 to 0.322 and valence from 0.244 to 0.368 on AVEC 2017 audio data versus standard BoW.