iPose: Instance-Aware 6D Pose Estimation of Partly Occluded Objects

TL;DR

This paper introduces iPose, a novel deep learning system for accurate 6D pose estimation of partly occluded objects from RGB and RGB-D images, using an instance-aware pipeline that decomposes the problem into simpler, sequential steps.

Contribution

The paper presents the first deep learning-based approach for 6D pose estimation of partly occluded objects from RGB input, utilizing a new instance-aware pipeline with multiple specialized steps.

Findings

Significantly outperforms previous methods on occluded object pose estimation

Effective in both RGB and RGB-D input scenarios

Demonstrates robustness to partial occlusions in complex scenes

Abstract

We address the task of 6D pose estimation of known rigid objects from single input images in scenarios where the objects are partly occluded. Recent RGB-D-based methods are robust to moderate degrees of occlusion. For RGB inputs, no previous method works well for partly occluded objects. Our main contribution is to present the first deep learning-based system that estimates accurate poses for partly occluded objects from RGB-D and RGB input. We achieve this with a new instance-aware pipeline that decomposes 6D object pose estimation into a sequence of simpler steps, where each step removes specific aspects of the problem. The first step localizes all known objects in the image using an instance segmentation network, and hence eliminates surrounding clutter and occluders. The second step densely maps pixels to 3D object surface positions, so called object coordinates, using an encoder-decoder network, and hence eliminates object appearance. The third, and final, step predicts the 6D pose using geometric optimization. We demonstrate that we significantly outperform the state-of-the-art for pose estimation of partly occluded objects for both RGB and RGB-D input.