swnet, paper(IEEE version), arxiv

Shared-Kernel Wavelet Networks for Near-Sensor Poisson Image Reconstruction, accepted by IEEE Sensors Journal

---

@ARTICLE{swnet,
  author={Gong, Yuanhao and Tang, Tan and Liu, Qianyan},
  journal={IEEE Sensors Journal}, 
  title={Shared-Kernel Wavelet Networks for Near-Sensor Poisson Image Reconstruction}, 
  year={2026},
  volume={},
  number={},
  pages={1-1},
  doi={10.1109/JSEN.2026.3702036}}

---

1) wavelet guided neural networks with shared-kernels

2) Laplacian Fields are sparse and statistically stable.

3) train a mini network for Poisson equation

4) thanks to the shared kernels, our method is mini

only 177 parameters

5) thanks to the data-driven, our method is more accurate