What to do with 2.000.000 Historical Press Photos? The Challenges and Opportunities of Applying a Scene Detection Algorithm to a Digitised Press Photo Collection

Abstract

In 1962, Dutch celebrity Ria Kuyken was attacked by a circus bear. Cees de Boer captured this moment, for which he was awarded both a World Press Photo and the Silver Camera (Zilveren Camera). Though this photo popularised Fotopersbureau De Boer, which Cees had founded in 1945, the importance of the collection lies in its scale. Approximately 2,000,000 photos taken of about 250,000 events in sixty years, accompanied by extensive metadata. Not only major nationwide events are represented, but also subjects of small scale, human interest, such as the shopkeeper around the corner. Our aim is not only the digitisation and publication of all 2,000,000 photo negatives of Fotopersbureau De Boer but also to explore how artificial intelligence can enrich this collection, benefiting both users of the archive and cultural historians studying historical photographs. One of our efforts focuses on scene detection, a method to detect the ‘scene’ represented in an image (Zhou et al, 2018). We will rely on transfer learning to adapt existing computer vision models to our collection and the needs of our users. Existing models can generate labels with high accuracy, however, these labels are ahistorical and more often than not irrelevant to our collection. We will label subsets of the images via crowdsourcing to train and improve existing models. As such, we can add labels relevant to our collection to the model, which are absent in existing models. In this paper, we will highlight the opportunities and challenges of applying artificial intelligence to a collection of historical photographs.