CAFE: Channel-Autoregressive Factorized Encoding for Robust Biosignal Spatial Super-Resolution

Hongjun Liu, Leyu Zhou, Zijianghao Yang, Rujun Han, Shitong Duan, Kuanjian Tang, Chao Yao · 2026

Read CAFE: Channel-Autoregressive Factorized Encoding for Robust Biosignal Spatial Super-Resolution by Hongjun Liu, Leyu Zhou, Zijianghao Yang, Rujun Han, Shitong Duan, Kuanjian Tang, Chao Yao (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/arxiv-2602-17011.txt.

Description

Hongjun Liu and co-authors propose CAFE, a channel-autoregressive factorized encoding scheme for robust biosignal spatial super-resolution. High-density biosignal recordings matter for neural decoding and clinical care, yet real setups often use low-density montages due to hardware and operational limits. That motivates spatial super-resolution from sparse data, but mixed dependencies in sparse, noisy measurements often lead to artifact spread and false links.

CAFE is a plug-and-play rollout scheme rebuilding the full montage in aligned stages. It first recovers nearby channels, then expands outward, using reliable local structure before introducing non-local interactions. During training, step-wise supervision over groups and teacher forcing with epoch-level scheduled sampling reduce exposure bias and allow parallel computation across steps. At test time, CAFE rolls out across groups, reusing any temporal backbone as predictor. Tests on four modalities and six datasets show it works on three backbones.

Catalog record

CREATORHongjun Liu, Leyu Zhou, Zijianghao Yang, Rujun Han, Shitong Duan, Kuanjian Tang, Chao Yao
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-17011

Browse the full library · About the Tower of Babel