Detection and classification of acoustic scenes and events: An IEEE AASP challenge

Giannoulis, Dimitrios; Benetos, Emmanouil; Stowell, Dan; Rossignol, Mathias; Lagrange, Mathieu; Plumbley, Mark D.

doi:10.1109/waspaa.2013.6701819

Cited by 159 publications

(137 citation statements)

References 8 publications

Supporting

Mentioning

130

Contrasting

Unclassified

Order By: Relevance

“…False negative denotes that the system fails to detect the active acoustic events in one segment. The length of the short segment is set to 100ms as in [27]. …”

Section: Evaluation Metricmentioning

confidence: 99%

Frame-Wise Dynamic Threshold Based Polyphonic Acoustic Event Detection

et al. 2017

View full text Add to dashboard Cite

Acoustic event detection, the determination of the acoustic event type and the localisation of the event, has been widely applied in many real-world applications. Many works adopt multi-label classification techniques to perform the polyphonic acoustic event detection with a global threshold to detect the active acoustic events. However, the global threshold has to be set manually and is highly dependent on the database being tested. To deal with this, we replaced the fixed threshold method with a frame-wise dynamic threshold approach in this paper. Two novel approaches, namely contour and regressor based dynamic threshold approaches are proposed in this work. Experimental results on the popular TUT Acoustic Scenes 2016 database of polyphonic events demonstrated the superior performance of the proposed approaches.

show abstract

“…False negative denotes that the system fails to detect the active acoustic events in one segment. The length of the short segment is set to 100ms as in [27]. …”

Section: Evaluation Metricmentioning

confidence: 99%

Frame-Wise Dynamic Threshold Based Polyphonic Acoustic Event Detection

et al. 2017

View full text Add to dashboard Cite

show abstract

“…For constructing the pre-extracted dictionary P (f |q, c, s), the IEEE DCASE Event Detection training dataset was used [7,1]. The dataset contains isolated sounds recorded in an office environment at Queen Mary University of London, and covers 16 event classes (s ∈ {1, ..., 16}): alert, clearing throat, cough, door slam, drawer, keyboard click, keys, door knock, laughter, mouse click, page turn, pen drop, phone, printer, speech, and switch.…”

Section: Training Datamentioning

confidence: 99%

“…PLCA with a single component was applied to each segment in order to extract a single sound state spectral template. For tuning system parameters for the polyphonic and monophonic datasets, the development datasets for the IEEE DCASE Office Synthetic and Office Live challenge [7] were used, respectively.…”

Section: Training Datamentioning

confidence: 99%

“…recent approach for event detection uses multilabel deep neural networks with spectral features as inputs [6]. In addition, as part of the IEEE AASP challenge on Detection and Classification of acoustic scenes and events (DCASE) [7], a baseline system was created using NMF with beta-divergence. Finally, also part of the DCASE challenge, Vuegen et al [8] proposed a system based on Gaussian mixture models (GMMs), with MFCCs as input features.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Detection of overlapping acoustic events using a temporally-constrained probabilistic model

Benetos

Lafay

Lagrange

et al. 2016

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

In this paper, a system for overlapping acoustic event detection is proposed, which models the temporal evolution of sound events. The system is based on probabilistic latent component analysis, supporting the use of a sound event dictionary where each exemplar consists of a succession of spectral templates. The temporal succession of the templates is controlled through event class-wise Hidden Markov Models (HMMs). As input time/frequency representation, the Equivalent Rectangular Bandwidth (ERB) spectrogram is used. Experiments are carried out on polyphonic datasets of office sounds generated using an acoustic scene simulator, as well as real and synthesized monophonic datasets for comparative purposes. Results show that the proposed system outperforms several state-of-the-art methods for overlapping acoustic event detection on the same task, using both frame-based and event-based metrics, and is robust to varying event density and noise levels.

show abstract

“…Acoustic events classification is an important topic in machine learning and data analysis [1][2][3]. Most of the traditional classifiers are built on a relatively small number of samples.…”

Section: Introductionmentioning

confidence: 99%

Acoustic Scene Classification on Large Dataset Using Sparse Auto-encoder Based Deep Neural Network

Tan¹

2017

Proceedings of the 2017 International Conference on Information Technology and Intelligent Manufacturing (ITIM 2017)

View full text Add to dashboard Cite

Abstract-In this paper we study the acoustic scene classification using a large dataset. The spectrogram of the large acoustic samples are extracted and applied with texture feature classification method. First, the acoustic scene database is built including various acoustic events. Second the image texture features on spectrogram are used to represent the acoustic samples. Third the auto-encoder is adopted to build a deep neural network classifier. Finally, we verified the proposed system on a large number of dataset and compared our results with traditional Gaussian mixture model and three-layer neural network. The experimental results show that the proposed method is effective and promising in big acoustic data classification.

show abstract

Detection and classification of acoustic scenes and events: An IEEE AASP challenge

Cited by 159 publications

References 8 publications

Frame-Wise Dynamic Threshold Based Polyphonic Acoustic Event Detection

Frame-Wise Dynamic Threshold Based Polyphonic Acoustic Event Detection

Detection of overlapping acoustic events using a temporally-constrained probabilistic model

Acoustic Scene Classification on Large Dataset Using Sparse Auto-encoder Based Deep Neural Network

Contact Info

Product

Resources

About