Massachusetts Institute of Technology (MIT) researchers have developed an algorithm that can identify which photos humans will find memorable, based on statistical data taken from a study of people remembering images. The researchers, led by MIT's Phillip Isola, found that the most memorable photos are those that contain people, followed by static indoor scenes and human-scale objects, while landscapes are not as memorable. After gathering the data, the researchers created memorability maps of each image by asking people to label all of the objects in the images. A computer model analyzed the maps to determine which objects make an image memorable. Then the researchers used machine-learning techniques to create a computational model that analyzed the images and their memorability rate, which enabled them to create an algorithm that can predict the memorability of images the computer has not previously analyzed.
No comments:
Post a Comment