Extraction and manipulation of wrinkles and spots for facial image synthesis

Extraction of wrinkles and spots in facial images is useful not only for face recognition, but also facial image synthesis. While many studies have focused on extraction of facial parts, such as eyes, a mouth, or a nose, only a few studies have attempted to extract wrinkles and spots. This paper proposes a novel method for extracting wrinkles and spots based on local analyses of the general shape properties of wrinkles and spots rather than applying global spatial filters as most previous methods do. Since the proposed method can extract wrinkles and spots separately, we can effectively manipulate them for facial image synthesis. We conducted an age evaluation experiment, confirming that effective facial image synthesis can be achieved when we manipulated not only facial parts, but also wrinkles and spots using our proposed method.