Ensemble of Exemplar-SVMs for Object Detection and Beyond.pdf

Ensemble of Exemplar-SVMs for Object Detection and Beyond Tomasz Malisiewicz Carnegie Mellon University Abhinav Gupta Carnegie Mellon University Alexei A. Efros Carnegie Mellon University Abstract This paper proposes a conceptually simple but surpris- ingly powerful method which combines the effectiveness of a discriminative object detector with the explicit correspon- dence offered by a nearest-neighbor approach. The method is based on training a separate linear SVM classifier for every exemplar in the training set. Each of these Exemplar- SVMs is thus defined by a single positive instance and mil- lions of negatives. While each detector is quite specific to its exemplar, we empirically observe that an ensemble of such Exemplar-SVMs offers surprisingly good generaliza- tion. Our performance on the PASCAL VOC detection task is on par with the much more complex latent part-based model of Felzenszwalb et al., at only a modest computa- tional cost increase. But the central benefit of our approach is that it creates an explicit association between each de- tection and a single training exemplar. Because most de- tections show good alignment to their associated exemplar, it is possible to transfer any available exemplar meta-data (segmentation, geometric structure, 3D model, etc.) directly onto the detections, which can then be used as part of over- all scene understanding. 1. Motivation A mere decade ago, automatically recognizing everyday objects in images (such as the bus in Figure 1) was thought to be an almost unsolvable task. Yet today, a number of methods can do just that with reasonable accuracy. But let us consider the output of a typical object detector – a rough bounding box around the object and a category label (Fig- ure 1 left). While this might be sufficient for a retrieval task (“find all buses in the database”), it seems rather lacking for any sort of deeper reasoning about the scene. How is the bus oriented? Is it a mini-bus or a double-decker? Which pi