ECHO: Environmental Sound Classification with Hierarchical Ontology-guided Semi-Supervised Learning

Résumé fourni par la source

Environment Sound Classification has been a well-studied research problem in the field of signal processing and till now more focus has been laid on fully supervised approaches. Recently, the focus has moved towards semi-supervised methods which concentrate on utilizing unlabeled data, and self-supervised methods which learn the intermediate representation through pretext tasks or contrastive learning. However, both approaches require a vast amount of unlabelled data to improve performance. In this work, we propose a novel framework called Environmental Sound Classification with Hierarchical Ontology-guided semi-supervised Learning (ECHO) that utilizes label ontology-based hierarchy to learn semantic representation by defining a novel pretext task. The model tries to predict coarse labels represented by the Large Language Model (LLM) based on ground truth label ontology, then further fine-tuned in a supervised way to predict the actual task. ECHO achieves a 1% to 8% accuracy improvement over baseline systems across UrbanSound8K, ESC-10, and ESC-50 datasets.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.