Multidimensional analyses of the noise impacts of COVID-19 lockdown.

As part of the Agence Nationale de Recherche Caractérisation des ENvironnements SonorEs urbains (Characterization of urban sound environments) project, a questionnaire was sent in January 2019 to households in a 1 km2 study area in the city of Lorient, France, to which about 318 responded. The main objective of this questionnaire was to collect information about the inhabitants' perception of the sound environments in their neighborhoods, streets, and dwellings. In the same study area, starting mid-2019, about 70 sensors were continuously positioned, and 15 of them were selected for testing sound source recognition models. The French lockdown due to the COVID-19 crisis occurred during the project, and the opportunity was taken to send a second questionnaire during April 2020. About 31 of the first 318 first survey respondents answered this second questionnaire. This unique longitudinal dataset, both physical and perceptual, allows the undertaking of an analysis from different perspectives of such a period. The analysis reveals the importance of integrating source recognition tools, soundscape observation protocol, in addition to physical level analysis, to accurately describe the changes in the sound environment.

Polyphonic training set synthesis improves self-supervised urban sound classification.

Machine listening systems for environmental acoustic monitoring face a shortage of expert annotations to be used as training data. To circumvent this issue, the emerging paradigm of self-supervised learning proposes to pre-train audio classifiers on a task whose ground truth is trivially available. Alternatively, training set synthesis consists in annotating a small corpus of acoustic events of interest, which are then automatically mixed at random to form a larger corpus of polyphonic scenes. Prior studies have considered these two paradigms in isolation but rarely ever in conjunction. Furthermore, the impact of data curation in training set synthesis remains unclear. To fill this gap in research, this article proposes a two-stage approach. In the self-supervised stage, we formulate a pretext task (Audio2Vec skip-gram inpainting) on unlabeled spectrograms from an acoustic sensor network. Then, in the supervised stage, we formulate a downstream task of multilabel urban sound classification on synthetic scenes. We find that training set synthesis benefits overall performance more than self-supervised learning. Interestingly, the geographical origin of the acoustic events in training set synthesis appears to have a decisive impact.

Brain-inspired model for early vocal learning and correspondence matching using free-energy optimization.

We propose a developmental model inspired by the cortico-basal system (CX-BG) for vocal learning in babies and for solving the correspondence mismatch problem they face when they hear unfamiliar voices, with different tones and pitches. This model is based on the neural architecture INFERNO standing for Iterative Free-Energy Optimization of Recurrent Neural Networks. Free-energy minimization is used for rapidly exploring, selecting and learning the optimal choices of actions to perform (eg sound production) in order to reproduce and control as accurately as possible the spike trains representing desired perceptions (eg sound categories). We detail in this paper the CX-BG system responsible for linking causally the sound and motor primitives at the order of a few milliseconds. Two experiments performed with a small and a large audio database show the capabilities of exploration, generalization and robustness to noise of our neural architecture in retrieving audio primitives during vocal learning and during acoustic matching with unheared voices (different genders and tones).

A Taxonomy Proposal for the Assessment of the Changes in Soundscape Resulting from the COVID-19 Lockdown.

Many countries around the world have chosen lockdown and restrictions on people's mobility as the main strategies to combat the COVID-19 pandemic. These actions have significantly affected environmental noise and modified urban soundscapes, opening up an unprecedented opportunity for research in the field. In order to enable these investigations to be carried out in a more harmonized and consistent manner, this paper makes a proposal for a set of indicators that will enable to address the challenge from a number of different approaches. It proposes a minimum set of basic energetic indicators, and the taxonomy that will allow their communication and reporting. In addition, an extended set of descriptors is outlined which better enables the application of more novel approaches to the evaluation of the effect of this new soundscape on people's subjective perception.

Gated spiking neural network using Iterative Free-Energy Optimization and rank-order coding for structure learning in memory sequences (INFERNO GATE).

We present a framework based on iterative free-energy optimization with spiking neural networks for modeling the fronto-striatal system (PFC-BG) for the generation and recall of audio memory sequences. In line with neuroimaging studies carried out in the PFC, we propose a genuine coding strategy using the gain-modulation mechanism to represent abstract sequences based solely on the rank and location of items within them. Based on this mechanism, we show that we can construct a repertoire of neurons sensitive to the temporal structure in sequences from which we can represent any novel sequences. Free-energy optimization is then used to explore and to retrieve the missing indices of the items in the correct order for executive control and compositionality. We show that the gain-modulation mechanism permits the network to be robust to variabilities and to have long-term dependencies as it implements a gated recurrent neural network. This model, called Inferno Gate, is an extension of the neural architecture Inferno standing for Iterative Free-Energy Optimization of Recurrent Neural Networks with Gating or Gain-modulation. In experiments performed with an audio database of ten thousand MFCC vectors, Inferno Gate is capable of encoding efficiently and retrieving chunks of fifty items length. We then discuss the potential of our network to model the features of working memory in the PFC-BG loop for structural learning, goal-direction and hierarchical reinforcement learning.

Kriging-based spatial interpolation from measurements for sound level mapping in urban areas.

Network-based sound monitoring systems are deployed in various cities over the world and mobile applications allowing participatory sensing are now common. Nevertheless, the sparseness of the collected measurements, either in space or in time, complicates the production of sound maps. This paper describes the results of a measurement campaign that has been conducted in order to test different spatial interpolation strategies for producing sound maps. Mobile measurements have been performed while walking multiple times in every street of the XIIIth district of Paris. By adaptively constructing a noise map on the basis of these measurements, the role of the density of observations and the performance of four different interpolation strategies is investigated. Ordinary and universal Kriging methods are assessed, as well as the effect of using an alternative definition of the distance between observation locations, which takes the topology of the road network into account. The results show that a high density of observation points is necessary to obtain an interpolated sound map close to the reference map.

An Efficient Audio Coding Scheme for Quantitative and Qualitative Large Scale Acoustic Monitoring Using the Sensor Grid Approach.

The spreading of urban areas and the growth of human population worldwide raise societal and environmental concerns. To better address these concerns, the monitoring of the acoustic environment in urban as well as rural or wilderness areas is an important matter. Building on the recent development of low cost hardware acoustic sensors, we propose in this paper to consider a sensor grid approach to tackle this issue. In this kind of approach, the crucial question is the nature of the data that are transmitted from the sensors to the processing and archival servers. To this end, we propose an efficient audio coding scheme based on third octave band spectral representation that allows: (1) the estimation of standard acoustic indicators; and (2) the recognition of acoustic events at state-of-the-art performance rate. The former is useful to provide quantitative information about the acoustic environment, while the latter is useful to gather qualitative information and build perceptually motivated indicators using for example the emergence of a given sound source. The coding scheme is also demonstrated to transmit spectrally encoded data that, reverted to the time domain using state-of-the-art techniques, are not intelligible, thus protecting the privacy of citizens.

Sound quality indicators for urban places in Paris cross-validated by Milan data.

A specific smartphone application was developed to collect perceptive and acoustic data in Paris. About 3400 questionnaires were analyzed, regarding the global sound environment characterization, the perceived loudness of some emergent sources and the presence time ratio of sources that do not emerge from the background. Sound pressure level was recorded each second from the mobile phone's microphone during a 10-min period. The aim of this study is to propose indicators of urban sound quality based on linear regressions with perceptive variables. A cross validation of the quality models extracted from Paris data was carried out by conducting the same survey in Milan. The proposed sound quality general model is correlated with the real perceived sound quality (72%). Another model without visual amenity and familiarity is 58% correlated with perceived sound quality. In order to improve the sound quality indicator, a site classification was performed by Kohonen's Artificial Neural Network algorithm, and seven specific class models were developed. These specific models attribute more importance on source events and are slightly closer to the individual data than the global model. In general, the Parisian models underestimate the sound quality of Milan environments assessed by Italian people.

Measurements of acoustic environments for urban soundscapes: choice of homogeneous periods, optimization of durations, and selection of indicators.

In order to minimize the duration of acoustic measurements and to characterize homogeneous areas from a temporal point of view, a series of six location measurements was carried out continuously during three months in Paris. Around fifty thousand samples of 5-min, 10-min, 15-min, 20-min, 30-min, and 1-h duration measurements were extracted for each location. Each sample is characterized by eleven energy indicators and ten event descriptors. In this paper, analysis of a crossroad location is detailed. Through hierarchical ascendant classification and artificial neural networks classification, it is shown that four homogeneous periods can be detected: two during the night, one during the day, and one transition corresponding either to the awakening or to the moment when the city falls asleep. 10-min measurements are necessary to discriminate these time periods at the crossroad location. At the end of the paper, a comparison with the other locations shows that minimum duration states in between 10 and 20 min. The homogeneous periods are connected to the human activities and depend on the location. Energy indicators such as LAeq, LA10, or LA90 and event indicators are necessary to characterize the different clusters.