polito.it
Politecnico di Torino (logo)

Study and implementation of a methodology for measuring and assessing the Lombard effect in eating establishments during COVID-19 pandemic

Giorgio Carzedda

Study and implementation of a methodology for measuring and assessing the Lombard effect in eating establishments during COVID-19 pandemic.

Rel. Arianna Astolfi, Marco Carlo Masoero, Giuseppina Emma Puglisi, Pasquale Bottalico. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Del Cinema E Dei Mezzi Di Comunicazione, 2021

[img]
Preview
PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (4MB) | Preview
Abstract:

Anthropic noise in highly attended environments has been observed and studied with increasing attention over the years. Very often the purpose of these studies has been finding several solutions to the acoustic discomfort linked to the noise generated by the presence of a large number of people, both indoors and outdoors. In particular, an attempt was made to analyze as thoroughly as possible the phenomenon whereby in very noisy environments a speaker unconsciously increases his vocal effort to always be understood by a listener: that is the Lombard effect. There are many researches in this regard, in which the relationship between noise and voice level has been examined in multiple conditions of both environments and social situations. Furthermore, starting from these observations there has been the attempt to formulate predictive models on noise produced in indoor environments by a large number of people speaking at the same time, in which the Lombard effect is one of the determining factors among others such as: room volume, reverberation time in unoccupied conditions, absorption coefficients, number of people present and ratio between the latter and the number of people who are actually speaking (group size). Predictive algorithms identified could find a new interpretation following the COVID-19 pandemic that began in 2020 and is still ongoing globally. In fact, in order to combat the disease and its extremely high contagiousness, the main objective of all the measures adopted was to avoid the gathering of people and to promote the social distance of at least 1 meter between each of them. In this perspective, a research project was launched at the Politecnico di Torino in which, using a prototype device called Speech and Sound SEMaphore (SEM), developed by the Politecnico itself to report excessive noise due to the aggregation of people, it is possible to monitor and indicate the presence of gatherings in a closed environment starting from the anthropic noise measured. This could be achieved by reversing the paradigm of predictive models discussed earlier in which the noise produced becomes a measured factor while the number of people present becomes the object of prediction. The subject of this thesis is to document the work I have done in this research project, in which my contribution has focused in particular on the development of a methodology useful for measurement, evaluation and parameterization of the Lombard effect. With this in mind, an experimental work was carried out in the field, during the course of 4 different days, carrying out noise measurements inside the staff canteen of the Politecnico di Torino. The measurements were made using both devices that recorded the individual voice levels of some participants in the research, and equipment that monitored the overall trend of the noise level within the environment. The data were subsequently processed through software by extrapolating the parameters useful for estimating the Lombard effect. Some considerations have been made on how to monitor and how to optimize the evaluation process, starting from a preliminary study on some cuts made on the recorded audio tracks and then automating the choice of these most representative segments. Finally, considerations were made on the results obtained and on the effectiveness of the methodology used, taking into account the exceptional circumstances in which it was applied (reduced canteen capacity, social distance).

Relatori: Arianna Astolfi, Marco Carlo Masoero, Giuseppina Emma Puglisi, Pasquale Bottalico
Anno accademico: 2020/21
Tipo di pubblicazione: Elettronica
Numero di pagine: 76
Soggetti:
Corso di laurea: Corso di laurea magistrale in Ingegneria Del Cinema E Dei Mezzi Di Comunicazione
Classe di laurea: Nuovo ordinamento > Laurea magistrale > LM-32 - INGEGNERIA INFORMATICA
Aziende collaboratrici: Politecnico di Torino
URI: http://webthesis.biblio.polito.it/id/eprint/19234
Modifica (riservato agli operatori) Modifica (riservato agli operatori)