A Two-Stage Transformer Framework for Sparse-Array Direction-of-Arrival Estimation via Correlation Vector Recovery
Wenchao He, Yiran Shi, Hongxi Zhao, Hongliang Zhu, Chunshan Bao · 2026
Read A Two-Stage Transformer Framework for Sparse-Array Direction-of-Arrival Estimation via Correlation Vector Recovery by Wenchao He, Yiran Shi, Hongxi Zhao, Hongliang Zhu, Chunshan Bao (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13210586.txt.
Description
Wenchao He and co-authors aim sound-source bearing with a two-stage transformer built for sparse, noisy sensor arrays. Finding where signals come from feeds radar, sonar, and microphone work. Missing sensors and few snapshots corrupt the correlation data that classic methods feed on.
Stage one rebuilds the full, clean correlation pattern from the ragged one, using masking-aware tokens and global context. That rebuilt pattern can become a tidy covariance matrix for classic back ends. Stage two reads the restored features and calls the bearings of several sources at once. Wide simulations across weak to fair signal levels show sturdier accuracy when snapshots run scarce. A classic ESPRIT back end also does better on the rebuilt data than on raw sparse feeds.
Catalog record
| CREATOR | Wenchao He, Yiran Shi, Hongxi Zhao, Hongliang Zhu, Chunshan Bao |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |