Sound Event Detection and Separation: A Benchmark on Desed Synthetic Soundscapes

Turpault, Nicolas; Serizel, Romain; Wisdom, Scott; Erdoğan, Hakan; Hershey, John R.; Fonseca, Eduardo; Seetharaman, Prem

doi:10.1109/icassp39728.2021.9414789

Cited by 23 publications

(16 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This experiment has already been proposed for DCASE 2019 and DCASE 2020 submissions [7,8]. Previous analyses showed that performance consistently drop when the onsets of the sound events are located towards the end of the clips [8]. Similar performance trends are obtained with general systems [15,19,20], while systems that have been adapted for scenario 1 [14,16,17] generally show attenuated performance drop towards the end of the clips.…”

Section: Impact Of Time Localization Of the Original Eventsupporting

confidence: 52%

“…We evaluate the systems using the evaluation sets described in Section 2.3.3. This experiment has already been proposed for DCASE 2019 and DCASE 2020 submissions [7,8]. Previous analyses showed that performance consistently drop when the onsets of the sound events are located towards the end of the clips [8].…”

Section: Impact Of Time Localization Of the Original Eventmentioning

confidence: 96%

“…These subsets are designed to study the sensibility of the SED segmentation to the sound event localization in time. A similar experiment has already been conducted for DCASE 2019 and DCASE 2020 Task 4 [7,8]. The main purpose here is to analyze whether the systems have improved on this particular aspect.…”

Section: Synthetic Set With Varying Onset Timementioning

confidence: 99%

“…This allows to generate custom synthetic clips to untangle some of the many problems faced by a SED system operating under real conditions. Since 2019, within DCASE Task 4, synthetic soundscapes have been designed to target some specific SED open problems and they have been proposed in the evaluation set to the challenge participants in order to obtain a benchmark of state-of-the-art SED systems on these specific problems [7,8].…”

Section: Introductionmentioning

confidence: 99%

“…During previous iterations of the challenge, we have observed that obtaining an accurate time segmentation was one of the most prominent challenges of the SED systems [8,9,10]. In 2021, we proposed to change the metric from an event-based F-score [11] to a polyphonic sound detection score (PSDS) [12].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes

Ronchini¹,

Serizel²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

This paper proposes a benchmark of submissions to Detection and Classification Acoustic Scene and Events 2021 Challenge (DCASE) Task 4 representing a sampling of the state-of-the-art in Sound Event Detection task. The submissions are evaluated according to the two polyphonic sound detection score scenarios proposed for the DCASE 2021 Challenge Task 4, which allow to make an analysis on whether submissions are designed to perform fine-grained temporal segmentation, coarse-grained temporal segmentation, or have been designed to be polyvalent on the scenarios proposed.We study the solutions proposed by participants to analyze their robustness to varying level target to non-target signal-to-noise ratio and to temporal localization of target sound events. A last experiment is proposed in order to study the impact of non-target events on systems outputs. Results show that systems adapted to provide coarse segmentation outputs are more robust to different target to non-target signal-to-noise ratio and, with the help of specific data augmentation methods, they are more robust to time localization of the original event. Results of the last experiment display that systems tend to spuriously predict short events when non-target events are present. This is particularly true for systems that are tailored to have a fine segmentation.

show abstract

Section: Impact Of Time Localization Of the Original Eventsupporting

confidence: 52%

Section: Impact Of Time Localization Of the Original Eventmentioning

confidence: 96%

Section: Synthetic Set With Varying Onset Timementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes

Ronchini¹,

Serizel²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Exploring current research trends in sound event detection: a systematic literature review

Mohmmad,

Sanampudi

2024

Multimed Tools Appl

View full text Add to dashboard Cite

Sound Event Detection for Human Safety and Security in Noisy Environments

et al. 2022

View full text Add to dashboard Cite

The objective of a sound event detector is to recognize anomalies in an audio clip and return their onset and offset. However, detecting sound events in noisy environments is a challenging task. This is due to the fact that in a real audio signal several sound sources co-exist and that the characteristics of polyphonic audio are different from isolated recordings. It is also necessary to consider the presence of noise (e.g. thermal and environmental). In this contribution, we present a sound anomaly detection system based on a fully convolutional network which exploits image spatial filtering and an Atrous Spatial Pyramid Pooling module. To cope with the lack of datasets specifically designed for sound event detection, a dataset for the specific application of noisy bus environments has been designed. The dataset has been obtained by mixing background audio files, recorded in a real environment, with anomalous events extracted from monophonic collections of labelled audios. The performances of the proposed system have been evaluated through segment-based metrics such as error rate, recall, and F1-Score. Moreover, robustness and precision have been evaluated through four different tests. The analysis of the results shows that the proposed sound event detector outperforms both state-of-the-art methods and general purpose deep learning-solutions.

show abstract

Sound Event Detection and Separation: A Benchmark on Desed Synthetic Soundscapes

Cited by 23 publications

References 18 publications

A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes

A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes

Exploring current research trends in sound event detection: a systematic literature review

Sound Event Detection for Human Safety and Security in Noisy Environments

Contact Info

Product

Resources

About