Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Bisot, Victor; Serizel, Romain; Essid, Slim; Richard, Gaël

doi:10.1109/taslp.2017.2690570

Cited by 92 publications

(79 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Our second baseline is a supervised NMF [20]. Our third baseline is a CNN optimized with categorical cross-entropy loss function, trained on the logarithmic conversion of the mel power of the input audios [21].…”

Section: A Baselinesmentioning

confidence: 99%

A hybrid approach with multi-channel i-vectors and convolutional neural networks for acoustic scene classification

Eghbal-zadeh

Lehner

Dorfer

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-In Acoustic Scene Classification (ASC) two major approaches have been followed . While one utilizes engineered features such as mel-frequency-cepstral-coefficients (MFCCs), the other uses learned features that are the outcome of an optimization algorithm. I-vectors are the result of a modeling technique that usually takes engineered features as input. It has been shown that standard MFCCs extracted from monaural audio signals lead to i-vectors that exhibit poor performance, especially on indoor acoustic scenes. At the same time, Convolutional Neural Networks (CNNs) are well known for their ability to learn features by optimizing their filters. They have been applied on ASC and have shown promising results. In this paper, we first propose a novel multi-channel i-vector extraction scheme for ASC, improving their performance on indoor and outdoor scenes. Second, we propose a CNN architecture that achieves promising ASC results. Further, we show that i-vectors and CNNs capture complementary information from acoustic scenes. Finally, we propose a hybrid system for ASC using multi-channel i-vectors and CNNs by utilizing a score fusion technique. Using our method, we participated in the ASC task of the DCASE-2016 challenge. Our hybrid approach achieved 1 st rank among 49 submissions, substantially improving the previous state of the art.

show abstract

Section: A Baselinesmentioning

confidence: 99%

A hybrid approach with multi-channel i-vectors and convolutional neural networks for acoustic scene classification

Eghbal-zadeh

Lehner

Dorfer

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

show abstract

“…TDL [13] has recently been applied with nonnegativity constraints to perform speech enhancement [15] or to acoustic scene classification, where temporally integrated projections are classified with multinomial logistic regression [14]. In this paper we extend the latter approach to the GNMF case.…”

Section: Task-driven Nmf Based Dictionary Learningmentioning

confidence: 99%

“…The general idea of nonnegative TDL or task-driven NMF (TNMF) is to unite the dictionary learning with NMF and the training of the classifier in a joint optimization problem [15,14]. Influenced by the classifier, the basis vectors are encouraged to explain the discriminative information in the data while keeping a low reconstruction cost.…”

Section: Task-driven Nmfmentioning

confidence: 99%

“…A supervised matrix factorization approach proposed recently and known as Task-driven Dictionary Learning (TDL) [13] allows for learning a dictionary jointly with a classifier, therefore enforcing the discriminative quality of the dictionary. This approach was later extended to nonnegative dictionaries and adapted to audio classification problem demonstrating significant performance improvement compared to unsupervised approaches [14].…”

Section: Introductionmentioning

confidence: 99%

“…The choice of the Euclidean distance renders the latter process more efficient and and allows to improve previous results be up to 9 % in some cases. In a first step, the nonnegative extension of the TDL [14,15] is applied to the dictionaries obtained with standard NMF or GNMF, in order to fine-tune the dictionaries. We then consider a taskdriven formulation for the GNMF method with speaker and session variability constraints [12].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Supervised group nonnegative matrix factorisation with similarity constraints and applications to speaker identification

Serizel

Bisot

Essid

et al. 2017

2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

This paper presents supervised feature learning approaches for speaker identification that rely on nonnegative matrix factorisation. Recent studies have shown that group nonnegative matrix factorisation and task-driven supervised dictionary learning can help performing effective feature learning for audio classification problems.This paper proposes to integrate a recent method that relies on group nonnegative matrix factorisation into a task-driven supervised framework for speaker identification. The goal is to capture both the speaker variability and the session variability while exploiting the discriminative learning aspect of the task-driven approach. Results on a subset of the ESTER corpus prove that the proposed approach can be competitive with I-vectors.

show abstract

Mixup-Based Acoustic Scene Classification Using Multi-channel Convolutional Neural Network

Feng

et al. 2018

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Audio scene classification, the problem of predicting class labels of audio scenes, has drawn lots of attention during the last several years. However, it remains challenging and falls short of accuracy and efficiency. Recently, Convolutional Neural Network (CNN)-based methods have achieved better performance with comparison to the traditional methods. Nevertheless, conventional single channel CNN may fail to consider the fact that additional cues may be embedded in the multi-channel recordings. In this paper, we explore the use of Multi-channel CNN for the classification task, which aims to extract features from different channels in an end-to-end manner. We conduct the evaluation compared with the conventional CNN and traditional Gaussian Mixture Model-based methods. Moreover, to improve the classification accuracy further, this paper explores the using of mixup method. In brief, mixup trains the neural network on linear combinations of pairs of the representation of audio scene examples and their labels. By employing the mixup approach for data augmentation, the novel model can provide higher prediction accuracy and robustness in contrast with previous models, while the generalization error can also be reduced on the evaluation data.

show abstract

Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Cited by 92 publications

References 45 publications

A hybrid approach with multi-channel i-vectors and convolutional neural networks for acoustic scene classification

A hybrid approach with multi-channel i-vectors and convolutional neural networks for acoustic scene classification

Supervised group nonnegative matrix factorisation with similarity constraints and applications to speaker identification

Mixup-Based Acoustic Scene Classification Using Multi-channel Convolutional Neural Network

Contact Info

Product

Resources

About