Please use this identifier to cite or link to this item:
|Title:||Urban sound tagging with spatiotemporal context||Authors:||Kwok, Li Long||Keywords:||Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Engineering::Computer science and engineering::Computing methodologies::Pattern recognition
|Issue Date:||2021||Publisher:||Nanyang Technological University||Source:||Kwok, L. L. (2021). Urban sound tagging with spatiotemporal context. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/149470||Project:||A3089-201||Abstract:||As Singapore continues to develop by building more residential and commercial infrastructure, noise caused by construction or traffic is unavoidable. To better control the noise, city planners will need to perform analysis to monitor the city's noise so that the citizens’ quality of life would not be affected. For the city planner to better understand and mitigate the noise pollution, an Urban Sound Tagging (UST) system was developed. The UST can be used to tag audio recordings of the city automatically. With the UST system, city planner can easily monitor the noise pollution in the city. In Detection and Classification of Acoustic Scenes and Events (DCASE) 2019 Task 5, many participants have developed UST that can tag on 10-second-long audio tracks to understand the city's noise profile. To boost the performance of the UST, DCASE 2020 Task 5 proposed a challenge named 'Urban Sound Tagging with Spatiotemporal Context'. This task was a follow up work from the previous year with additional metadata such as Spatiotemporal (STC) metadata. The task aims to investigate whether the STC metadata would aid in the performance of the UST system. With the inspiration from DCASE 2020 Task 5, this project investigates the proposed models from DCASE 2020 Task 5 to observe whether there was a performance boost by using STC metadata. Therefore, an experiment will be conducted by training different metadata and the performance will be evaluated. In addition, the adaptability of the models will be evaluated as well by introducing a new dataset named the Sound of Singapore (SGS) dataset to the model for tagging of soundscape that was recorded in Singapore.||URI:||https://hdl.handle.net/10356/149470||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Updated on May 28, 2022
Updated on May 28, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.