Domain-incremental audio classification using domain-specific experts and prototype classifier
For researchers in continual learning for audio classification, this work provides a practical solution to catastrophic forgetting, though it is an incremental improvement over existing methods.
The authors tackle domain-incremental audio classification without access to past or future domain data. Their system, using frozen experts and a prototype classifier, achieves 78.15% micro and 77.03% macro accuracy on the development set, outperforming individual backbones.
This technical report presents submission systems for Task 7(domain-incremental audio classification) of the DCASE 2026 Challenge. The main obstacle is that, the system is unable to access to past or future domain's data at once. We approached domain-incremental learning (DIL) as a frozen-feature replay problem. At each incremental stage, one or two compact experts are trained and then kept fixed; at the final stage, the penultimate features from all frozen experts are concatenated and used to train a lightweight per-class prototype classifier solely on cached features. This design prevents catastrophic forgetting by preserving each expert models at inference. To retain earlier-domain knowledge without storing raw audio, some experts were trained with DeepInversion-based generative replay. A cross-stage regression imputer was trained to fill the expert feature slots that did not yet exist at an ealier stage. We submit four fully DIL-compliant systems: three systems based on diverse frozen five-expert backbones and their cross-stack ensemble achieving 78.15% micro / 77.03% macro on the development set, outperforming every individual backbone on both evaluations.