Probabilistic Segmentation for Robust Field of View Estimation
R. Spencer Hallyburton, David Hunt, Yiwei He, Judy He, Miroslav Pajic · 2025
Read Probabilistic Segmentation for Robust Field of View Estimation by R. Spencer Hallyburton, David Hunt, Yiwei He, Judy He, Miroslav Pajic (2025) 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-2503-07375.txt.
Description
Spencer Hallyburton, David Hunt, Yiwei He, Judy He, and Miroslav Pajic build tools that map what a self-driving car's sensors can see. Their paper from twenty twenty-five notes that safe sensor fusion needs a sound field of view map. No autonomy-grade gauge of such maps had been made. The team adapts classic graphics methods to build the first such estimators.
They also made the first data sets with true field of view labels. Tests then showed a catch: the classic estimators break under sensor attacks. The fix is a learned segmentation model that reads view traits. Monte Carlo dropout gives it a sense of its own doubt. Odd confidence maps flag attacks. Broad tests show strong attack resistance and good spread to new scenes. Trade studies say it can run in real time.
Catalog record
| CREATOR | R. Spencer Hallyburton, David Hunt, Yiwei He, Judy He, Miroslav Pajic |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-07375