A Conceptual Framework for Learning to Listen by Reward: Curiosity-Driven Search for Novel Sources

MCML Authors

Andreas Triantafyllopoulos

Dr.

→ Group Björn Schuller
 Health Informatics

Björn Schuller

Prof. Dr.

Core PI

Health Informatics

Abstract

Reinforcement learning is a powerful learning paradigm that has spearheaded progress in numerous domains. Its core promise lies in learning through high-level goals without the need for granular labels. However, it still remains elusive in the realm of audio, where it has received substantially less attention than in computer vision or other domains. The key question remains: how can agents learn to listen purely via reward-driven exploration? In this contribution, we present an overview of previous attempts and a new conceptual framework for learning to listen by reward. Our approach depends on the continuous search for novel sound sources. We formulate our framework, discuss open technical challenges, and present a first proof-of-concept implementation that showcases the feasibility of our approach.

misc TST+26

---

Preprint

May. 2026

---

Authors

A. Triantafyllopoulos • J. Šťastný • A. Terpinas • T. Liu • Y. Wang • B. W. Schuller

---

Links

arXiv

---

Research Area

 B3 | Multimodal Perception