Image ID by Neural Network

Are you interested in the power of machine learning? What do you think the odds are that a simple app can identify the dominant object in a photograph of your choosing? This app demonstrates the CoreML software framework using the Resnet50 image classification model (https://arxiv.org/abs/1512.03385). To use this app, simply select a photo from your library or take a new picture with your camera and see how the neural network classification works! The Resnet50 model authors state that their approach "detects the dominant objects present in an image from a set of 1000 categories such as trees, animals, food, vehicles, person etc. The top-5 error from the original publication is 7.8%." You will discover that the model isn't perfect and not all categories are represented, but it delivers some interesting and surprisingly accurate predictions for many photo subjects!