An i‐vector representation based on bottleneck (BN) features is presented for language identification (LID). In the proposed system, the BN features are extracted from a deep neural network, which can effectively mine the contextual information embedded in speech frames. The i‐vector representation of each utterance is then obtained by applying a total variability approach on the BN features. The resulting performance of LID has been significantly improved with the proposed BN feature based i‐vector representation. Compared with the state‐of‐the‐art techniques, the equal error rate is relatively reduced by about 40% on the National Institute of Standards and Technology (NIST) 2009 evaluation sets.

A key problem in spoken language identification (LID) is to design effective representations which are specific to language information. For example, in recent years, representations based on both phonotactic and acoustic features have proven their effectiveness for LID. Although advances in machine learning have led to significant improvements, LID performance is still lacking, especially for short duration speech utterances. With the hypothesis that language information is weak and represented only latently in speech, and is largely dependent on the statistical properties of the speech content, existing representations may be insufficient. Furthermore they may be susceptible to the variations caused by different speakers, specific content of the speech segments, and background noise. To address this, we propose using Deep Bottleneck Features (DBF) for spoken LID, motivated by the success of Deep Neural Networks (DNN) in speech recognition. We show that DBFs can form a low-dimensional compact representation of the original inputs with a powerful descriptive and discriminative capability. To evaluate the effectiveness of this, we design two acoustic models, termed DBF-TV and parallel DBF-TV (PDBF-TV), using a DBF based i-vector representation for each speech utterance. Results on NIST language recognition evaluation 2009 (LRE09) show significant improvements over state-of-the-art systems. By fusing the output of phonotactic and acoustic approaches, we achieve an EER of 1.08%, 1.89% and 7.01% for 30 s, 10 s and 3 s test utterances respectively. Furthermore, various DBF configurations have been extensively evaluated, and an optimal system proposed.

This paper explores the use of three different two-dimensional time–frequency features for audio event classification with deep neural network back-end classifiers. The evaluations use spectrogram, cochleogram and constant-Q transform-based images for classification of 50 classes of audio events in varying levels of acoustic background noise, revealing interesting performance patterns with respect to noise level, feature image type and classifier. Evidence is obtained that two well-performing features, the spectrogram and cochleogram, make use of information that is potentially complementary in the input features. Feature fusion is thus explored for each pair of features, as well as for all tested features. Results indicate that a fusion of spectrogram and cochleogram information is particularly beneficial, yielding an impressive 50-class accuracy of over \(96\%\) in 0 dB SNR and exceeding \(99\%\) accuracy in 10 dB SNR and above. Meanwhile, the cochleogram image feature is found to perform well in extreme noise cases of \(-\,5\) dB and \(-\,10\) dB SNR.

This paper presents an improved mean teacher (MT) based method for large-scale weakly labeled semi-supervised sound event detection (SED), by focusing on learning a better student model. Two main improvements are proposed based on the authors’ previous perturbation based MT method. Firstly, an event-aware module is de-signed to allow multiple branches with different kernel sizes to be fused via an attention mechanism. By inserting this module after the convolutional layer, each neuron can adaptively adjust its receptive field to suit different sound events. Secondly, instead of using the teacher model to provide a consistency cost term, we propose using a stochastic inference of unlabeled examples to generate high quality pseudo-targets by averaging multiple predictions from the perturbed student model. MixUp of both labeled and unlabeled data is further exploited to improve the effectiveness of student model. Finally, the teacher model can be obtained via exponential moving average (EMA) of the student model, which generates final predictions for SED during inference. Experiments on the DCASE2018 task4 dataset demonstrate the ability of the proposed method. Specifically, an F1-score of 42.1% is achieved, significantly outperforming the 32.4% achieved by the winning system, or the 39.3% by the previous perturbation based method.