Recent progress in single-pixel imaging (SPI) has exhibited remarkable performance using deep neural networks, e.g., convolutional neural networks (CNNs) and vision Transformers (ViTs). Nonetheless, it is challenging for existing methods to well model object i... ...
