Efficient detection for Spatially Local Coding

Abstract

In this paper, we present an efficient detector for the Spatially Local Coding (SLC) object model. SLC is a recent, high performing object classifier that has yet to be applied in a detection (object localization) setting. SLC uses features that jointly code for both appearance and location, making it difficult to apply the existing approaches to efficient detection. We design an approximate Hough transform for the SLC model that uses a cascade of thresholds followed by gradient descent to achieve efficiency as well as accurate localization. We evaluate the resulting detector on the Daimler Monocular Pedestrian dataset.