A Hybrid Time and Time-frequency Domain Implicit Neural Representation for Acoustic Fields

Ge Z, Li L, Qu T. A Hybrid Time and Time-frequency Domain Implicit Neural Representation for Acoustic Fields. In: Express Paper. Audio Engineering Society; 2024. Paper 196. Available from: https://aes.org/publications/elibrary-page/?id=22542

@inproceedings{Ge2024_22542,
  author = {Ge, Zhongshu and Li, Liang and Qu, Tianshu},
  title = {{A Hybrid Time and Time-frequency Domain Implicit Neural Representation for Acoustic Fields}},
  note = {Paper 196},
  year = {2024},
  month = jun,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=22542}
}

TY  - CPAPER
TI  - A Hybrid Time and Time-frequency Domain Implicit Neural Representation for Acoustic Fields
AU  - Ge, Zhongshu
AU  - Li, Liang
AU  - Qu, Tianshu
M1  - Paper 196
PY  - 2024
DA  - 2024/06/06
UR  - https://aes.org/publications/elibrary-page/?id=22542
PB  - Audio Engineering Society
LA  - en
AB  - Creating an immersive scene relies on detailed spatial sound. Traditional methods, using probe points for impulse responses, need lots of storage. Meanwhile, geometry-based simulations struggle with complex sound effects. Now, neural-based methods are improving accuracy and slashing storage needs. In our study, we propose a hybrid time and time-frequency domain strategy to model the time series of Ambisonic acoustic fields. The networks excels in generating high-fidelity time-domain impulse responses at arbitrary source-recceiver positions by learning a continuous representation of the acoustic field. Our experimental results demonstrate that the proposed model outperforms baseline methods in various aspects of sound representation and rendering for different source-recceiver positions.
ER  -