Residual Learning for Neural Ambisonics Encoders
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915755849678848 |
|---|---|
| author | Deppisch, Thomas Gao, Yang Mittal, Manan Stahl, Benjamin Hold, Christoph Alon, David Ben-Hur, Zamir |
| author_facet | Deppisch, Thomas Gao, Yang Mittal, Manan Stahl, Benjamin Hold, Christoph Alon, David Ben-Hur, Zamir |
| contents | Emerging wearable devices such as smartglasses and extended reality headsets demand high-quality spatial audio capture from compact, head-worn microphone arrays. Ambisonics provides a device-agnostic spatial audio representation by mapping array signals to spherical harmonic (SH) coefficients. In practice, however, accurate encoding remains challenging. While traditional linear encoders are signal-independent and robust, they amplify low-frequency noise and suffer from high-frequency spatial aliasing. On the other hand, neural network approaches can outperform linear encoders but they often assume idealized microphones and may perform inconsistently in real-world scenarios. To leverage their complementary strengths, we introduce a residual-learning framework that refines a linear encoder with corrections from a neural network. Using measured array transfer functions from smartglasses, we compare a UNet-based encoder from the literature with a new recurrent attention model. Our analysis reveals that both neural encoders only consistently outperform the linear baseline when integrated within the residual learning framework. In the residual configuration, both neural models achieve consistent and significant improvements across all tested metrics for in-domain data and moderate gains for out-of-domain data. Yet, coherence analysis indicates that all neural encoder configurations continue to struggle with directionally accurate high-frequency encoding. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18322 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Residual Learning for Neural Ambisonics Encoders Deppisch, Thomas Gao, Yang Mittal, Manan Stahl, Benjamin Hold, Christoph Alon, David Ben-Hur, Zamir Audio and Speech Processing Sound Emerging wearable devices such as smartglasses and extended reality headsets demand high-quality spatial audio capture from compact, head-worn microphone arrays. Ambisonics provides a device-agnostic spatial audio representation by mapping array signals to spherical harmonic (SH) coefficients. In practice, however, accurate encoding remains challenging. While traditional linear encoders are signal-independent and robust, they amplify low-frequency noise and suffer from high-frequency spatial aliasing. On the other hand, neural network approaches can outperform linear encoders but they often assume idealized microphones and may perform inconsistently in real-world scenarios. To leverage their complementary strengths, we introduce a residual-learning framework that refines a linear encoder with corrections from a neural network. Using measured array transfer functions from smartglasses, we compare a UNet-based encoder from the literature with a new recurrent attention model. Our analysis reveals that both neural encoders only consistently outperform the linear baseline when integrated within the residual learning framework. In the residual configuration, both neural models achieve consistent and significant improvements across all tested metrics for in-domain data and moderate gains for out-of-domain data. Yet, coherence analysis indicates that all neural encoder configurations continue to struggle with directionally accurate high-frequency encoding. |
| title | Residual Learning for Neural Ambisonics Encoders |
| topic | Audio and Speech Processing Sound |
| url | https://arxiv.org/abs/2601.18322 |