Viewpoint Adaptation for Person Detection

An object detector performs suboptimally when applied to image data taken from a viewpoint different from the one with which it was trained. In this paper, we present a viewpoint adaptation algorithm that allows a trained single-view person detector to be adapted to a new, distinct viewpoint. We first illustrate how a feature space transformation can be inferred from a known homography between the source and target viewpoints. Second, we show that a variety of trained classifiers can be modified to behave as if that transformation were applied to each testing instance. The proposed algorithm is evaluated on a new synthetic multi-view dataset as well as images from the PETS 2007 and CAVIAR datasets, yielding substantial performance improvements when adapting single-view person detectors to new viewpoints while increasing the detector frame rate. This work has the potential to improve person detection performance for cameras at non-standard viewpoints while simplifying data collection and feature extraction.