Geometric Correspondence Fields: Learned Differentiable Rendering for 3D Pose Refinement in the Wild

TL;DR

This paper introduces a novel differentiable rendering method that uses learned geometric correspondence fields to refine 3D object poses in real-world images, significantly improving accuracy over previous techniques.

Contribution

It proposes a new feature space comparison for pose refinement and a learned differentiable renderer that predicts pixel-level correspondences for precise 3D pose alignment.

Findings

Achieves up to 55% relative improvement on Pix3D dataset

Outperforms state-of-the-art refinement methods in multiple metrics

Introduces a learned backward pass for differentiable rendering

Abstract

We present a novel 3D pose refinement approach based on differentiable rendering for objects of arbitrary categories in the wild. In contrast to previous methods, we make two main contributions: First, instead of comparing real-world images and synthetic renderings in the RGB or mask space, we compare them in a feature space optimized for 3D pose refinement. Second, we introduce a novel differentiable renderer that learns to approximate the rasterization backward pass from data instead of relying on a hand-crafted algorithm. For this purpose, we predict deep cross-domain correspondences between RGB images and 3D model renderings in the form of what we call geometric correspondence fields. These correspondence fields serve as pixel-level gradients which are analytically propagated backward through the rendering pipeline to perform a gradient-based optimization directly on the 3D pose. In this way, we precisely align 3D models to objects in RGB images which results in significantly improved 3D pose estimates. We evaluate our approach on the challenging Pix3D dataset and achieve up to 55% relative improvement compared to state-of-the-art refinement methods in multiple metrics.