This is the repository for the "Images to live with" tool.
The tool automatically picks from a selection of 14436 images randomly chosen from 8 different datasets: ("Tiny ImageNet", "COIL-100", "Labelled Faces in the Wild", "Leaf", "Open Image", "Stanford Dogs", "Tencent", and "VisualGenome"). It uses a simple Node.js server with the pdfkit library to generate PDFs containing the random images from datasets.
For more details on the process, read the article "Living with images from large-scale data sets: A critical pedagogy for scaling down", on the Photographies journal.