A Cold Diffusion Approach for Percussive Dereverberation

Dimos Makris, Maximos Kaliakatsos-Papakostas

Keywords: cold diffusion, drums dereverberation, audio enhancement, generative models, music production

Domains: Machine Learning Theory, Other

TL;DR: A novel cold diffusion approach for percussive dereverberation on stereo drums in the Real/Imaginary STFT domain.

External Link: https://arxiv.org/abs/2605.10256

Abstract: Most recent advances in audio dereverberation focus almost exclusively on speech, leaving percussive and drum signals largely unexplored despite their importance in music production. Percussive dereverberation poses distinct challenges due to sharp transients and dense temporal structure. In this work, we propose a cold diffusion framework for dereverberating stereo drum stems (downmixes), modeling reverberation as a deterministic degradation process that progressively transforms anechoic signals into reverberant ones. We investigate two reverse-process parameterizations—Direct (next-state) and a Delta-normalized residual (velocity-style) prediction—and implement the framework using both a UNet and a diffusion Transformer backbone. The models are trained and evaluated on curated datasets comprising both acoustic and electronic drum recordings, with reverberation generated using a combination of synthetic and real room impulse responses. Extensive experiments on in-domain and fully out-of-domain test sets demonstrate that the proposed method consistently outperforms strong score-based and conditional diffusion baselines, evaluated using signal-based and perceptual metrics tailored to percussive audio. This paper is accepted for the 2026 IEEE World Congress on Computational Intelligence, IJCNN Track, 21-26 June 2026, Maastricht, the Netherlands.

Submission Number: 152