MonStereo: When Monocular and Stereo Meet at the Tail of 3D Human Localization

TL;DR

This paper introduces MonStereo, a unified framework combining monocular and stereo cues for improved 3D human localization, especially in occluded or distant scenarios, with a new evaluation metric for human localization.

Contribution

It presents a novel joint learning approach that leverages both monocular and stereo information, addressing their individual limitations for 3D human localization.

Findings

Effective handling of occlusions and distant pedestrians

Improved accuracy over existing methods on KITTI dataset

Proposed a practical 3D localization metric for humans

Abstract

Monocular and stereo visions are cost-effective solutions for 3D human localization in the context of self-driving cars or social robots. However, they are usually developed independently and have their respective strengths and limitations. We propose a novel unified learning framework that leverages the strengths of both monocular and stereo cues for 3D human localization. Our method jointly (i) associates humans in left-right images, (ii) deals with occluded and distant cases in stereo settings by relying on the robustness of monocular cues, and (iii) tackles the intrinsic ambiguity of monocular perspective projection by exploiting prior knowledge of the human height distribution. We specifically evaluate outliers as well as challenging instances, such as occluded and far-away pedestrians, by analyzing the entire error distribution and by estimating calibrated confidence intervals. Finally, we critically review the official KITTI 3D metrics and propose a practical 3D localization metric tailored for humans.