Acoustic context recognition using local binary pattern codebooks

Battaglino, Daniele; Lepauloux, Ludovick; Pilati, Laurent; Evans, Nicholas

doi:10.1109/waspaa.2015.7336886

Cited by 17 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They aim at characterizing the spectro-temporal variations of acoustic events occurring in a scene by computing the gradient of pixels in time-frequency images. In addition to HOG, other spectrogram image-based features have been proposed such as the Subband Power Distribution (SPD) [13] or Local Binary Patterns [29].…”

Section: Related Workmentioning

confidence: 99%

Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Bisot

Serizel

Essid

et al. 2017

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Bisot

Serizel

Essid

et al. 2017

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

“…Image features such as LBP, Sub-band Power Distribution (SPD) [28] and HOG have been exploited in ASC. The CQT and spectrogram have been widely used as TFRs to integrate with the feature learning method or to be coupled with hand-crafted features such as HOG and LBP [8], [29], [16], [30], [9]. The texture features extracted from the TFRs capture the time and frequency discriminative patterns of the audio structure.…”

Section: Previous Workmentioning

confidence: 99%

“…Ye et al [31] incorporated the statistics of local pixel values of the spectrogram into the LBP. In [30], LBP features were extracted from the spectrogram and the bag-of-features model was applied to generate LBP-Codebook features. Recently, [32] proposed the application of LBP to capture the temporal dynamic features of MFCC's representation.…”

Section: Previous Workmentioning

confidence: 99%

Spectrotemporal Analysis Using Local Binary Pattern Variants for Acoustic Scene Classification

Abidin

Togneri

Sohel

2018

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

In this paper we present an approach for acoustic scene classification, which aggregates spectral and temporal features. We do this by proposing the first use of the variable-Q transform (VQT) to generate the time-frequency representation for acoustic scene classification. The VQT provides finer control over the resolution compared to the constant-Q transform (CQT) or STFT and can be tuned to better capture acoustic scene information. We then adopt a variant of the local binary pattern (LBP), the Adjacent Evaluation Completed LBP (AECLBP), which is better suited to extracting features from acoustic time-frequency images. Our results yield a 5.2% improvement on the DCASE 2016 dataset compared to the application of standard CQT with LBP. Fusing our proposed AECLBP with HOG features we achieve a classification accuracy of 85.5% which outperforms one of the top performing systems.

show abstract

“…Acoustic scene analysis, has received a lot of research attention recently [1,2], which aims to recognize acoustic environments [3,4]. It finds applications in many audio devices, such as cars, robots, context-aware mobile devices, and intelligent monitoring systems.…”

Section: Introductionmentioning

confidence: 99%

An Unsupervised Deep Learning System for Acoustic Scene Analysis

2020

View full text Add to dashboard Cite

Acoustic scene analysis has attracted a lot of attention recently. Existing methods are mostly supervised, which requires well-predefined acoustic scene categories and accurate labels. In practice, there exists a large amount of unlabeled audio data, but labeling large-scale data is not only costly but also time-consuming. Unsupervised acoustic scene analysis on the other hand does not require manual labeling but is known to have significantly lower performance and therefore has not been well explored. In this paper, a new unsupervised method based on deep auto-encoder networks and spectral clustering is proposed. It first extracts a bottleneck feature from the original acoustic feature of audio clips by an auto-encoder network, and then employs spectral clustering to further reduce the noise and unrelated information in the bottleneck feature. Finally, it conducts hierarchical clustering on the low-dimensional output of the spectral clustering. To fully utilize the spatial information of stereo audio, we further apply the binaural representation and conduct joint clustering on that. To the best of our knowledge, this is the first time that a binaural representation is being used in unsupervised learning. Experimental results show that the proposed method outperforms the state-of-the-art competing methods.

show abstract

Acoustic context recognition using local binary pattern codebooks

Cited by 17 publications

References 12 publications

Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Feature Learning With Matrix Factorization Applied to Acoustic Scene Classification

Spectrotemporal Analysis Using Local Binary Pattern Variants for Acoustic Scene Classification

An Unsupervised Deep Learning System for Acoustic Scene Analysis

Contact Info

Product

Resources

About