Skip to Main content Skip to Navigation
Conference papers

Task-Aware Separation for the DCASE 2020 Task 4 Sound Event Detection and Separation Challenge

Abstract : Source Separation is often used as a pre-processing step in many signal-processing tasks. In this work we propose a novel approach for combined Source Separation and Sound Event Detection in which a Source Separation algorithm is used to enhance the Sound Even-Detection back-end performance. In particular, we present a permutation-invariant training scheme for optimizing the Source Separation system directly with the back-end Sound Event Detection objective without requiring joint training or fine-tuning of the two systems. We show that such an approach has significant advantages over the more standard approach of training the Source Separation system separately using only a Source Separation based objective such as Scale-Invariant Signal-To-Distortion Ratio. On the 2020 Detection and Classification of Acoustic Scenes and Events Task 4 Challenge our proposed approach is able to outperform the baseline source separation system by more than one percent in event-based macro F1 score on the development set with significantly less computational requirements.
Complete list of metadatas

Cited literature [30 references]  Display  Hide  Download

https://hal.inria.fr/hal-02962907
Contributor : Manuel Pariente <>
Submitted on : Friday, October 9, 2020 - 3:45:30 PM
Last modification on : Wednesday, October 14, 2020 - 4:21:59 AM

File

Task_aware_separation_DCASE202...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02962907, version 1

Collections

Citation

Samuele Cornell, Michel Olvera, Manuel Pariente, Giovanni Pepe, Emanuele Principi, et al.. Task-Aware Separation for the DCASE 2020 Task 4 Sound Event Detection and Separation Challenge. DCASE 2020 - 5th Workshop on Detection and Classification of Acoustic Scenes and Events, Nov 2020, Virtual, Japan. ⟨hal-02962907⟩

Share

Metrics

Record views

89

Files downloads

61