Incremental Learning of Acoustic Scenes and Sound Events

TL;DR

This paper introduces an incremental learning approach for acoustic scene classification and audio tagging using a CNN, addressing catastrophic forgetting with independent learning and knowledge distillation, and demonstrating effective performance on TUT datasets.

Contribution

It presents a novel incremental learning method combining independent learning and knowledge distillation to mitigate forgetting in acoustic scene and sound event tasks.

Findings

Achieved 94.0% accuracy on ASC task

Obtained 54.4% F1 score on AT task

Minimal performance decrease on previous tasks

Abstract

In this paper, we propose a method for incremental learning of two distinct tasks over time: acoustic scene classification (ASC) and audio tagging (AT). We use a simple convolutional neural network (CNN) model as an incremental learner to solve the tasks. Generally, incremental learning methods catastrophically forget the previous task when sequentially trained on a new task. To alleviate this problem, we propose independent learning and knowledge distillation (KD) between the timesteps in learning. Experiments are performed on TUT 2016/2017 dataset, containing 4 acoustic scene classes and 25 sound event classes. The proposed incremental learner first solves the ASC task with an accuracy of 94.0%. Next, it learns to solve the AT task with an F1 score of 54.4%. At the same time, its performance on the previous ASC task decreases only by 5.1 percentage points due to the additional learning of the AT task.