Synthetic Occlusion Augmentation with Volumetric Heatmaps for the 2018 ECCV PoseTrack Challenge on 3D Human Pose Estimation

TL;DR

This paper introduces a 3D human pose estimation method that uses volumetric heatmaps and synthetic occlusion augmentation, achieving top results in the ECCV PoseTrack Challenge and surpassing state-of-the-art on Human3.6M without extra datasets.

Contribution

The paper presents a novel augmentation technique with synthetic occlusions for improved 3D pose estimation accuracy.

Findings

Achieved first place in the ECCV PoseTrack Challenge 2018.

Surpassed state-of-the-art on Human3.6M without additional datasets.

Effective use of volumetric heatmaps and synthetic occlusion augmentation.

Abstract

In this paper we present our winning entry at the 2018 ECCV PoseTrack Challenge on 3D human pose estimation. Using a fully-convolutional backbone architecture, we obtain volumetric heatmaps per body joint, which we convert to coordinates using soft-argmax. Absolute person center depth is estimated by a 1D heatmap prediction head. The coordinates are back-projected to 3D camera space, where we minimize the L1 loss. Key to our good results is the training data augmentation with randomly placed occluders from the Pascal VOC dataset. In addition to reaching first place in the Challenge, our method also surpasses the state-of-the-art on the full Human3.6M benchmark among methods that use no additional pose datasets in training. Code for applying synthetic occlusions is availabe at https://github.com/isarandi/synthetic-occlusion.