This paper presents a new model of object classes which incorporates appearance and shape information jointly. Modeling objects appearance by distributions of visual words has recently proven successful. Here appearance based models are augmented by capturing the spatial arrangement of visual words. Compact spatial modeling without loss of discrimination is achieved through the introduction of adaptive vector quantized correlograms, which we call correlatons. Efﬁciency is further improved by means of integral images. The robustness of our new models to geometric transformations, severe occlusions and missing information is also demonstrated. The accuracy of discrimination of the proposed models is assessed with respect to existing databases with large numbers of object classes viewed under general conditions, and shown to outperform appearance-only models.