Using Context to Recognize People in Consumer Images

摘要: Recognizing people in images is one of the foremost challenges computer vision. It important to remember that consumer photography has a highly social aspect. The photographer captures not random fashion, but rather or document meaningful events her life. Understanding necessitates context each person an image considered. Context includes information related scene surrounding person, camera such as location and capture time, describes interactions between people. goal this paper provide with same intuition humans would use for analyzing Fortunately, than relying on lifetime experience, can often be modeled large amounts publicly available data. Probabilistic graph models machine learning are used model relationship principled manner.