Abstract

We propose a novel and efficient method for generic arbitrary- view ob ject class detection and localization. In contrast to existing single- view and multi-view methods using complicated mechanisms for relating the structural information in different parts of the ob jects or different view- points, we aim at representing the structural information in their true 3D locations. Uncalibrated multi-view images from a hand-held camera are used to reconstruct the 3D visual word models in the training stage. In the testing stage, beyond bounding boxes, our method can automatically de- termine the locations and outlines of multiple ob jects in the test image with occlusion handling, and can accurately estimate both the intrinsic and ex- trinsic camera parameters in an optimized way. With exemplar models, our method can also handle shape deformation for intra-class variance. To han- dle large data sets from models, we propose several speedup techniques to make the prediction efficient. Experimental results obtained based on some standard data sets demonstrate the effectiveness of the proposed approach.