InSPyReNet is a deep learning model for salient object detection that revisits image pyramid structures to handle high-resolution inputs. It is built with PyTorch and achieves state-of-the-art results on standard benchmarks like HKU-IS, DUTS-TE, PASCAL-S, ECSSD, and DUT-OMRON. The method employs a dichotomous image segmentation approach and is designed for accurate foreground object detection in images.