TY - RPRT TI - Training DeepFilterNet with Accurate Room Acoustic Simulations Improves Single-Channel Speech Enhancement AU - Alessia Milo AU - Georg Götz AU - Steinar Guðjónsson AU - Daniel Gert Nielsen AU - Jesper Pedersen AU - Finnur Pind PY - 2026 UR - https://arxiv.org/abs/2608.20971 ID - 2608.20971 ER -