English

PISE: Physics-Anchored Semantically-Enhanced Deep Computational Ghost Imaging for Robust Low-Bandwidth Machine Perception

Computer Vision and Pattern Recognition 2026-03-16 v2 Image and Video Processing

Abstract

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.

Cite

@article{arxiv.2601.12551,
  title  = {PISE: Physics-Anchored Semantically-Enhanced Deep Computational Ghost Imaging for Robust Low-Bandwidth Machine Perception},
  author = {Tong Wu},
  journal= {arXiv preprint arXiv:2601.12551},
  year   = {2026}
}

Comments

4 pages, 4 figures, 4 tables. Refined version with updated references and formatting improvements

R2 v1 2026-07-01T09:09:43.630Z