Joint analysis of acoustic scenes and sound events based on multitask learning with dynamic weight adaptation

Abstract

Acoustic scene classification (ASC) and sound event detection (SED) are major topics in environmental sound analysis. Considering that acoustic scenes and sound events are closely related to each other, the joint analysis of acoustic scenes and sound events using multitask learning (MTL)-based neural networks was proposed in some previous works. Conventional methods train MTL-based models using a linear combination of ASC and SED loss functions with constant weights. However, the performance of conventional MTL-based methods depends strongly on the weights of the ASC and SED losses, and it is difficult to determine the appropriate balance between the constant weights of the losses of MTL of ASC and SED. In this paper, we thus propose dynamic weight adaptation methods for MTL of ASC and SED based on dynamic weight average (DWA) and multi-focal loss (MFL) to adjust the learning weights automatically. By comparing the two methods, we then clarify how the dynamic adaptation of the loss weights, rather than specific methods of DWA and MFL, generally benefits the joint analysis of ASC and SED based on MTL. Moreover, we investigate how the training of the joint ASC and SED model dynamically progresses and disclose how the loss weights affect their performance.