Multi-Scale Gist Feature Representation for Building Recognition

Building recognition is a relatively specific recognition task in object recognition, which is a challenging task since it encounters rotation, illumination changes, occlusion, etc. But human can recognize the gist of a novel image in a single glance despite of its complexity. Inspired by this human vision characteristic, we describe a new building recognition model, called multi-scale gist feature representation model,which captures a holistic and low-dimensional representation of the structure of a building image. To evaluate the performance of our proposed model, experiments were carried out on the Sheffield buildings database, compared with the existing works:(a) the visual gist based building recognition model (VGBR); (b) the hierarchical building recognition model (HBR). The results show that the proposed model is effective and robust.