A Parametric Dual-Channel Audio Coding via Learned Time-Frequency Masking

Wang Y, Qian Y, Huang Q, Qu T. A Parametric Dual-Channel Audio Coding via Learned Time-Frequency Masking. In: Convention Paper. Audio Engineering Society; 2026.pp. 10294:1-10. Available from: https://aes.org/publications/elibrary-page/?id=23191

@inproceedings{Wang2026_23191,
  author = {Wang, Yihan and Qian, Yufan and Huang, Qingbo and Qu, Tianshu},
  title = {{A Parametric Dual-Channel Audio Coding via Learned Time-Frequency Masking}},
  pages = {10294:1-10},
  url = {https://aes.org/publications/elibrary-page/?id=23191},
  year = {2026},
  month = may,
  publisher = {Audio Engineering Society},
  address = {Copenhagen Denmark}
}

TY  - CPAPER
TI  - A Parametric Dual-Channel Audio Coding via Learned Time-Frequency Masking
AU  - Wang, Yihan
AU  - Qian, Yufan
AU  - Huang, Qingbo
AU  - Qu, Tianshu
SP  - 1
EP  - 10
M1  - Paper 10294
PY  - 2026
DA  - 2026/05/28
UR  - https://aes.org/publications/elibrary-page/?id=23191
PB  - Audio Engineering Society
CY  - Copenhagen Denmark
LA  - en
AB  - While Neural Audio Codecs (NAC) have revolutionized monaural audio compression, achieving high-fidelity dual-channel coding at low bitrates remains a significant challenge. Existing approaches often rely on naive independent channel quantization, leading to phase incoherence, or entangled latent modeling, which sacrifices spatial precision for spectral energy. This paper proposes a novel dual-channel coding framework based on contentspatial disentanglement. Reframing spatial reconstruction as an informed source separation task, our architecture synergizes a frozen, pre-trained DAC encoder for robust mono content preservation with a parameter-efficient side information encoder that predicts fine-grained time-frequency masks. To ensure precise spatial imaging, we introduce explicit physical constraints into the end-to-end training. Experimental results indicate that at low bitrates of 9 and 11 kbps, the proposed method outperforms state-of-the-art dual-mono neural baselines and industry standards in both objective spatial metrics and subjective MUSHRA evaluations.
ER  -