Probabilistic Vehicle Reconstruction Using a Multi-Task CNN

TL;DR

This paper introduces a probabilistic 3D vehicle reconstruction method from stereo images that uses a multi-task CNN to predict distributions for vehicle pose and shape, improving robustness and accuracy.

Contribution

The paper presents a novel multi-task CNN that outputs probability distributions for vehicle keypoints, edges, and orientation, integrated into a probabilistic framework for improved 3D reconstruction.

Findings

Achieves state-of-the-art results on KITTI benchmark

Reduces sensitivity to illumination and keypoint localization errors

Demonstrates effectiveness on new Stereo-Vehicle dataset

Abstract

The retrieval of the 3D pose and shape of objects from images is an ill-posed problem. A common way to object reconstruction is to match entities such as keypoints, edges, or contours of a deformable 3D model, used as shape prior, to their corresponding entities inferred from the image. However, such approaches are highly sensitive to model initialisation, imprecise keypoint localisations and/or illumination conditions. In this paper, we present a probabilistic approach for shape-aware 3D vehicle reconstruction from stereo images that leverages the outputs of a novel multi-task CNN. Specifically, we train a CNN that outputs probability distributions for the vehicle's orientation and for both, vehicle keypoints and wireframe edges. Together with 3D stereo information we integrate the predicted distributions into a common probabilistic framework. We believe that the CNN-based detection of wireframe edges reduces the sensitivity to illumination conditions and object contrast and that using the raw probability maps instead of inferring keypoint positions reduces the sensitivity to keypoint localisation errors. We show that our method achieves state-of-the-art results, evaluating our method on the challenging KITTI benchmark and on our own new 'Stereo-Vehicle' dataset.