@misc{indiciae689903a6a3f5,
  title = {Time-Frequency Weighted Losses for Phoneme Reconstruction in DNN-Based Speech Enhancement},
  author = {Nasser-Eddine Monir and Paul Magron and Romain Serizel},
  year = {2026},
  url = {https://arxiv.org/abs/2606.21635},
  note = {Source identifier: 2606.21635}
}