The paper presents a system for recognizing human faces from single images in a large database, where each person has only one image. The system addresses the challenge of image variation due to position, size, expression, and pose by extracting concise face descriptions in the form of *image graphs*. These graphs represent fiducial points on the face using sets of wavelet components (*jets*). The extraction of image graphs is based on a novel approach called *bunch graph*, which is constructed from a small set of sample image graphs. Recognition is achieved by comparing image graphs. The system is evaluated on the FERET and Bochum databases, including recognition across different poses. The results show high recognition rates, especially for frontal views, and the system outperforms previous methods in terms of speed and accuracy.The paper presents a system for recognizing human faces from single images in a large database, where each person has only one image. The system addresses the challenge of image variation due to position, size, expression, and pose by extracting concise face descriptions in the form of *image graphs*. These graphs represent fiducial points on the face using sets of wavelet components (*jets*). The extraction of image graphs is based on a novel approach called *bunch graph*, which is constructed from a small set of sample image graphs. Recognition is achieved by comparing image graphs. The system is evaluated on the FERET and Bochum databases, including recognition across different poses. The results show high recognition rates, especially for frontal views, and the system outperforms previous methods in terms of speed and accuracy.