PISE: Physics-Anchored Semantically-Enhanced Deep Computational Ghost Imaging for Robust Low-Bandwidth Machine Perception
arXiv:2601.125514.1
Predicted impact top 99% in CV · last 90 daysOriginality Incremental advance
AI Analysis
This work addresses robust machine perception for edge devices with limited bandwidth, representing an incremental improvement in computational ghost imaging.
The paper tackled the problem of low-bandwidth edge perception by proposing PISE, a physics-informed deep ghost imaging framework, which improved classification accuracy by 2.57% and reduced variance by 9x at 5% sampling.
We propose PISE, a physics-informed deep ghost imaging framework for low-bandwidth edge perception. By combining adjoint operator initialization with semantic guidance, PISE improves classification accuracy by 2.57% and reduces variance by 9x at 5% sampling.