I've been exploring the Royal Photographic Society (RPS) collection with data visualisation (see here). After segmenting the data by photographic technique and themes, I was interested to see if I could trace visual characteristics across the photograph collection; this is not possible by relying solely on the information in our catalogue. The collection is partly digitised and, at the time of writing, I've been working with around 1,200 items with an image and date. One way to explore an image collection visually is to use a machine-learning model to extract feature vectors from the images. Feature vectors are (multi-dimensional) numerical measures that describe the visual characteristics of an image.