The model takes color images which have been converted to grayscale and converts them back, trying to minimize an AI-measured “perceptual distance” between its colorized versions and the originals. While ImageNet likely contains more white people than people of color, there’s another source of bias as well. In fact, the model wasn’t trained on historical photos at all! It was trained on the ImageNet dataset, put together by researchers at Stanford in 2009 with Flickr photos. But it isn’t perfect the colors are dulled.Ĭontrary to what you might think, this problem isn’t happening because there are more white people in some historical photo dataset, or because those photos have more beige colors. It uses a sophisticated image generating model called a Generative Adversarial Network, and Antic’s algorithm for training it works pretty well, which results in a reliable image colorizer that produces realistic-looking images. The colorization algorithm in question is Jason Antic’s DeOldify, which you can try out on. Rather than write lots of rules, we use machine learning to build a statistical model from data about which colors most likely occur. The woman’s dress could be blue, but it could be another color, and there’s no information in the grayscale pixels to indicate which, so an algorithm has to take an informed guess. Beyond the obvious problems with lightening a woman’s skin, photographs have a lot of power over how we imagine and feel about the world, and seeing the past with dulled color makes it look dead.Ĭolorization is hard for computers because it is *ill-posed,* meaning that there are multiple color images which are equally “correct” given a grayscale root version. As an AI researcher interested in history, I find this issue troubling.
0 kommentar(er)
