AI RESEARCH
Learning Binary Sampling Patterns for Single-Pixel Imaging using Bilevel Optimisation
arXiv CS.LG
•
ArXi:2508.19068v2 Announce Type: replace-cross Single-Pixel Imaging (SPI) enables the reconstruction of objects using a single detector through sequential illuminations with structured light patterns. The choice of illumination patterns is critical, particularly in highly undersampled regimes, where it directly determines reconstruction quality and acquisition speed. Instead of relying on handcrafted or fixed patterns, we propose to learn task-specific patterns directly from data. Practical SPI hardware only s binary patterns, making binary pattern design a necessary consideration.