Acoustic event detection in real life recordings

By Annamaria Mesaros, Toni Heittola, Antti J. Eronen, Tuomas Virtanen. 2010

331 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

FSD50K: An Open Dataset of Human-Labeled Sound Events (2020, methodology): “Early stage works in SER relied on feature engineering approaches using standard machine learning classifiers such as support vector machines [14], Gaussian mixture models [15] or matrix factorization techniques [16].”
CNN architectures for large-scale audio classification (2016, methodology): “Historically, AED has been addressed with features such as MFCCs and classifiers based on GMMs, HMMs, NMF, or SVMs [8, 9, 10, 11].”
PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition (2019, methodology): “[29] A.”
Detection and Classification of Acoustic Scenes and Events (2015, background): “The majority of work in event detection treats the sound signal as monophonic, with only one event detectable at a time [19], [20].”
Sound Event Localization and Detection of Overlapping Sources Using Convolutional Recurrent Neural Networks (2018, background): “Some of the classiﬁers include Gaussian mixture model (GMM) - hidden Markov model (HMM) [27], fully connected (FC) neural networks [28], recurrent neural networks (RNN) [29– 32], and convolutional neural networks (CNN) [33, 34].”
Recurrent neural networks for polyphonic sound event detection in real life recordings (2016, methodology): “Initial approaches to polyphonic SED include traditional methods for speech recognition, such as the use of mel frequency cepstral coefficients (MFCCs) as features, with Gaussian mixture models (GMMs) combined with hidden Markov models (HMMs) [5, 6].”
Polyphonic sound event detection using multi label deep neural networks (2015, methodology): “The same database has been previously used in different experiments on sound event detection [1], [5], [6].”
Convolutional Recurrent Neural Networks for Polyphonic Sound Event Detection (2017, unknown): “Gaussian mixture model (GMM) - Hidden Markov model (HMM) based modeling— an established method that has been widely used in automatic speech recognition—has been proposed to model individual sound events with Gaussian mixtures and detect each event through HMM states using Viterbi algorithm [12], [1”
Detection and Classification of Acoustic Scenes and Events: Outcome of the DCASE 2016 Challenge (2018, methodology): “On the acoustic modeling aspect, methods range from classical statistical models like hidden Markov models (HMMs) [8], Gaussian mixture models (GMMs) [9] or support vector machines (SVMs) [11], to more recently developed methods using deep learning that have high computational complexity in training”
DCASE2017 Challenge Setup: Tasks, Datasets and Baseline System (2017, methodology): “The most used features for sound event detection are mel-scale representations, namely cepstral coefficients or log energies [12, 13, 14], and they are used with various machine learning methods, including HMMs [12], non-negative matrix factorization (NMF) [15, 16], random forests [11], and DNNs [14”
Metrics for Polyphonic Sound Event Detection (2016, background): “Sound event classiﬁcation in its simplest form requires assigning an event class label to each test audio, as illustrated in Figure 1A. Classiﬁcation is performed on audio containing isolated sound events [10,13,14] or containing a target sound event and additional overlapping sounds [23].”