Machine learning in acoustics: theory and applications

Abstract

Acoustic data provide scientific and engineering insights in fields ranging from biology and communications to ocean and Earth science. We survey the recent advances and transformative potential of machine learning (ML), including deep learning, in the field of acoustics. ML is a broad family of techniques, which are often based in statistics, for automatically detecting and utilizing patterns in data. Relative to conventional acoustics and signal processing, ML is data-driven. Given sufficient training data, ML can discover complex relationships between features and desired labels or actions, or between features themselves. With large volumes of training data, ML can discover models describing complex acoustic phenomena such as human speech and reverberation. ML in acoustics is rapidly developing with compelling results and significant future promise. We first introduce ML, then highlight ML developments in four acoustics research areas: source localization in speech processing, source localization in ocean acoustics, bioacoustics, and environmental sounds in everyday scenes.

Published with free access in Journal of the Acoustical Society of America, 27 Nov. 2019

References in corpus (25)

• Adam: A Method for Stochastic Optimization
• Scikit-learn: Machine Learning in Python
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• R-FCN: Object Detection via Region-based Fully Convolutional Networks
• Deep Learning for Audio Signal Processing
• Understanding Deep Convolutional Networks
• MatConvNet - Convolutional Neural Networks for MATLAB
• Convolutional Recurrent Neural Networks for Polyphonic Sound Event Detection
• Sound Event Localization and Detection of Overlapping Sources Using Convolutional Recurrent Neural Networks
• Online EM Algorithm for Latent Data Models
• Acoustic Scene Classification
• Multi-Speaker DOA Estimation Using Deep Convolutional Networks Trained with Noise Signals
• Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning
• Broadband DOA estimation using Convolutional neural networks trained with noise signals
• SoundNet: Learning Sound Representations from Unlabeled Video
• Source localization in an ocean waveguide using supervised machine learning
• Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions
• Audio Event Detection using Weakly Labeled Data
• Deep-learning source localization using multi-frequency magnitude-only data
• Sparse Bayesian learning with uncertainty models and multiple dictionaries
• Sound source ranging using a feed-forward neural network with fitting-based early stopping
• Multiple-Speaker Localization Based on Direct-Path Features and Likelihood Maximization with Spatial Sparsity Regularization
• Local Identification of Overcomplete Dictionaries
• Multichannel Online Dereverberation based on Spectral Magnitude Inverse Filtering

Cited by in corpus (33)

• Deep Networks for Direction-of-Arrival Estimation in Low SNR
• A Survey of Sound Source Localization with Deep Learning Methods
• Improving aircraft performance using machine learning: a review
• A Review of Machine Learning Methods Applied to Structural Dynamics and Vibroacoustic
• Seabed classification using physics-based modeling and machine learning
• Deep embedded clustering of coral reef bioacoustics
• Tensor-based Basis Function Learning for Three-dimensional Sound Speed Fields
• Ultrathin, high-speed, all-optical photoacoustic endomicroscopy probe for guiding minimally invasive surgery
• Semi-supervised source localization with deep generative modeling
• Training a U-Net based on a random mode-coupling matrix model to recover acoustic interference striations
• DeepLofargram: A Deep Learning based Fluctuating Dim Frequency Line Detection and Recovery
• BeamLearning: an end-to-end Deep Learning approach for the angular localization of sound sources using raw multichannel acoustic pressure data
• Complex data labeling with deep learning methods: Lessons from fisheries acoustics
• Underwater Sound Speed Profile Construction: A Review
• ElectrodeNet -- A Deep Learning Based Sound Coding Strategy for Cochlear Implants
• Classifying acoustic cavitation with machine learning trained on multiple physical models
• Automatic source localization and spectra generation from sparse beamforming maps
• Audio scene monitoring using redundant ad-hoc microphone array networks
• Acoustic Structure Inverse Design and Optimization Using Deep Learning
• Machine Learning in Acoustics: A Review and Open-Source Repository
• Towards Robust Data-Driven Underwater Acoustic Localization: A Deep CNN Solution with Performance Guarantees for Model Mismatch
• Mean absorption estimation from room impulse responses using virtually supervised learning
• Model reduction in acoustic inversion by artificial neural network
• Experimental Results of Underwater Sound Speed Profile Inversion by Few-shot Multi-task Learning
• A Review on Sound Source Localization in Robotics: Focusing on Deep Learning Methods
• Expert Decision Support System for aeroacoustic source type identification using clustering
• STNet: Prediction of Underwater Sound Speed Profiles with An Advanced Semi-Transformer Neural Network
• A t-distribution based operator for enhancing out of distribution robustness of neural network classifiers
• Neural network for multi-exponential sound energy decay analysis
• Noise signal as input data in self-organized neural networks
• Synthesis of Soundfields through Irregular Loudspeaker Arrays Based on Convolutional Neural Networks
• Prediction of acoustic field in 1-D uniform duct with varying mean flow and temperature using neural networks
• An experiment on an automated literature survey of data-driven speech enhancement methods