Image Recognition
..“Now, using digital photography of pottery, computers have hundreds by archaeologists who physically grouped broken pieces of pottery in a fraction of the time and in a more consistent way. In 2016, he and professor of anthropology, Chris Downham, began investigating the possibility of using a computer to accurately classify pieces of pottery, called debris, into known pottery types. If Pawlowicz and Downum can use CNN to identify dog breeds, consumer-favorite products, etc., why not apply this approach to the analysis of ancient pottery? Next, we hired four top Southwestern pottery experts to identify the pottery design types for all huts and create a hut “training set” where machines could learn. Finally, they trained the machine to learn the types of pottery by focusing on the pottery specimens that archaeologists agreed to.