How Good are Local Features for Classes of Geometric Objects

Recent work in object categorization often uses local image descriptors such as SIFT to learn and detect object categories. As such descriptors explicitly code local appearance they have shown impressive results on objects with sufficient local appearance statistics. However, many important object classes such as tools, cups and other man-made artifacts seem to require features that capture the respective shape and geometric layout of those object classes. Therefore this paper compares, on a novel data collection of 10 geometric object classes, various shape-based features with more appearance based descriptors such as SIFT. The analysis includes a direct comparison of feature statistics as well as the results within standard recognition frameworks. The results suggest that there are indeed differences between shape- based and more appearance-based features but that those differences do not always conform with what one might expect.