Multi-modal 3D Human Pose Estimation with 2D Weak Supervision in Autonomous Driving

TL;DR

This paper introduces a multi-modal approach for 3D human pose estimation in autonomous vehicles that leverages 2D image labels as weak supervision, combining LiDAR and camera data to improve accuracy.

Contribution

It presents one of the first methods to use 2D weak supervision with multi-modal data for 3D HPE in autonomous driving, addressing data scarcity issues.

Findings

22% improvement over camera-only 2D HPE baseline

6% improvement over LiDAR-only model

Effective use of multi-modal data with auxiliary segmentation

Abstract

3D human pose estimation (HPE) in autonomous vehicles (AV) differs from other use cases in many factors, including the 3D resolution and range of data, absence of dense depth maps, failure modes for LiDAR, relative location between the camera and LiDAR, and a high bar for estimation accuracy. Data collected for other use cases (such as virtual reality, gaming, and animation) may therefore not be usable for AV applications. This necessitates the collection and annotation of a large amount of 3D data for HPE in AV, which is time-consuming and expensive. In this paper, we propose one of the first approaches to alleviate this problem in the AV setting. Specifically, we propose a multi-modal approach which uses 2D labels on RGB images as weak supervision to perform 3D HPE. The proposed multi-modal architecture incorporates LiDAR and camera inputs with an auxiliary segmentation branch. On the Waymo Open Dataset, our approach achieves a 22% relative improvement over camera-only 2D HPE baseline, and 6% improvement over LiDAR-only model. Finally, careful ablation studies and parts based analysis illustrate the advantages of each of our contributions.