TY  - RPRT
TI  - Beyond Oversmoothing: Evaluating DDPM and MSE for Scalable Speech Synthesis in ASR
AU  - Christoph Minixhofer
AU  - Ondrej Klejch
AU  - Peter Bell
PY  - 2024
UR  - https://arxiv.org/abs/2410.12279
ID  - 2410.12279
ER  -