Resilient Multiple Choice Learning: A learned scoring scheme with application to audio scene analysis
This work addresses multimodal regression challenges in audio scene analysis, offering an incremental improvement over existing Multiple Choice Learning methods.
The paper tackles the problem of multimodal density estimation in regression by introducing Resilient Multiple Choice Learning (rMCL), which uses a learned scoring scheme based on Voronoi tessellations to preserve prediction diversity, and demonstrates its application to sound source localization with empirical validation.
We introduce Resilient Multiple Choice Learning (rMCL), an extension of the MCL approach for conditional distribution estimation in regression settings where multiple targets may be sampled for each training input. Multiple Choice Learning is a simple framework to tackle multimodal density estimation, using the Winner-Takes-All (WTA) loss for a set of hypotheses. In regression settings, the existing MCL variants focus on merging the hypotheses, thereby eventually sacrificing the diversity of the predictions. In contrast, our method relies on a novel learned scoring scheme underpinned by a mathematical framework based on Voronoi tessellations of the output space, from which we can derive a probabilistic interpretation. After empirically validating rMCL with experiments on synthetic data, we further assess its merits on the sound source localization problem, demonstrating its practical usefulness and the relevance of its interpretation.