Occlusion-Aware Object Localization, Segmentation and Pose Estimation

TL;DR

This paper introduces an occlusion-aware learning method for object localization, segmentation, and pose estimation that outperforms previous approaches by effectively handling partial occlusions using higher order potentials and a specialized loss function.

Contribution

It proposes a novel occlusion-aware model with higher order potentials and an efficient loss function, significantly improving segmentation and localization accuracy under occlusion.

Findings

Achieves 13.52% segmentation error on CMU Kitchen dataset

Improves localization performance by 16.13% over state-of-the-art

Enables robust 3D pose estimation from a single image despite occlusions

Abstract

We present a learning approach for localization and segmentation of objects in an image in a manner that is robust to partial occlusion. Our algorithm produces a bounding box around the full extent of the object and labels pixels in the interior that belong to the object. Like existing segmentation aware detection approaches, we learn an appearance model of the object and consider regions that do not fit this model as potential occlusions. However, in addition to the established use of pairwise potentials for encouraging local consistency, we use higher order potentials which capture information at the level of im- age segments. We also propose an efficient loss function that targets both localization and segmentation performance. Our algorithm achieves 13.52% segmentation error and 0.81 area under the false-positive per image vs. recall curve on average over the challenging CMU Kitchen Occlusion Dataset. This is a 42.44% decrease in segmentation error and a 16.13% increase in localization performance compared to the state-of-the-art. Finally, we show that the visibility labelling produced by our algorithm can make full 3D pose estimation from a single image robust to occlusion.