Acoustic Scene Classification: Classifying environments from the sounds they produce

By Daniele Barchiesi, Dimitrios Giannoulis, Dan Stowell, Mark D. Plumbley. IEEE Signal Processing Magazine, 2014

434 citations recorded by the source corpus.

In this article, we present an account of the state of the art in acoustic scene classification (ASC), the task of classifying environments from the sounds they produce. Starting from a historical review of previous research in this area, we define a general framework for ASC and present different implementations of its components. We then describe a range of different algorithms submitted for a data challenge that was held to provide a general and fair benchmark for ASC techniques. The data set recorded for this purpose is presented along with the performance metrics that are used to evaluate the algorithms and statistical significance tests to compare the submitted methods.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/0ae21e1bddb58e31ede3fc036ec63c8f79bd6533

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

A multi-device dataset for urban acoustic scene classification (2018, methodology): “In 2013, the acoustic scene classiﬁcation task used a development dataset consisting of 10 acoustic scenes each with 10 examples of 30 s, and an evaluation dataset of the same size [1, 4].”
Acoustic Scene Classification by Ensembling Gradient Boosting Machine and Convolutional Neural Networks (2017, background): “, Acoustic Scene Classification (ASC), that can be defined as the task of associating a label to an audio stream thereby identifying the particular context or environment where the audio stream was generated [1].”
Detection and Classification of Acoustic Scenes and Events (2015, methodology): “Figure is adapted from [64].”
Detection and Classification of Acoustic Scenes and Events: Outcome of the DCASE 2016 Challenge (2018, methodology): “is presented in [7], presenting in detail the approaches submitted for DCASE 2013.”
Chime-home: A dataset for sound source recognition in a domestic environment (2015, background): “SSR closely relates to acoustic scene classiﬁcation (ASC) [7], which aims at assigning a single semantic label to an audio recording, describing the environment which produced the acoustic scene.”
CNN-Based Learnable Gammatone Filterbank and Equal-Loudness Normalization for Environmental Sound Classification (2020, background): “E NVIRONMENTAL sound classiﬁcation (ESC) is the task of classifying an environment according to its ambient sound [1].”
Environmental sound classification with convolutional neural networks (2015, background): “The introduction of deep learning techniques in this context has slowly begun in the last two years.”
DCASE2017 Challenge Setup: Tasks, Datasets and Baseline System (2017, methodology): “A review of the features and classifiers used for it is presented in [7], with features including the well-known mel-frequency cepstral coefficients [2, 8] or more specialized features such as histograms of sound events [9] or histogram of gradients learned from time-frequency representations [10], ”
WavCaps: A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research (2023, background): “This has resulted in the development of algorithms that can detect and identify sound events and acoustic scenes [4]–[7].”