Class-Incremental Learning for Multi-Label Audio Classification

TL;DR

This paper introduces a class-incremental learning method for multi-label audio classification that effectively learns new sound classes while retaining knowledge of previous classes, demonstrated by consistent performance on a 50-class dataset.

Contribution

It proposes a novel incremental learning approach using cosine similarity and KL divergence-based distillation losses for multi-label audio classification.

Findings

Achieved an average F1-score of 40.9% over five phases.

Maintained a performance degradation of only 0.7 percentage points.

Performed well on a dataset with 50 sound classes.

Abstract

In this paper, we propose a method for class-incremental learning of potentially overlapping sounds for solving a sequence of multi-label audio classification tasks. We design an incremental learner that learns new classes independently of the old classes. To preserve knowledge about the old classes, we propose a cosine similarity-based distillation loss that minimizes discrepancy in the feature representations of subsequent learners, and use it along with a Kullback-Leibler divergence-based distillation loss that minimizes discrepancy in their respective outputs. Experiments are performed on a dataset with 50 sound classes, with an initial classification task containing 30 base classes and 4 incremental phases of 5 classes each. After each phase, the system is tested for multi-label classification with the entire set of classes learned so far. The proposed method obtains an average F1-score of 40.9% over the five phases, ranging from 45.2% in phase 0 on 30 classes, to 36.3% in phase 4 on 50 classes. Average performance degradation over incremental phases is only 0.7 percentage points from the initial F1-score of 45.2%.