Fast 3D Pose Refinement with RGB Images

TL;DR

This paper introduces a CNN-based system that refines coarse 3D pose estimates from simpler algorithms using RGB images, achieving high precision with minimal training data, suitable for resource-constrained robots.

Contribution

The paper presents a novel, efficient 3D pose refinement method that improves accuracy using RGB images and minimal training, reducing reliance on complex neural networks.

Findings

Refines coarse 3D poses to high precision

Operates with minimal training data

Suitable for resource-limited robotic systems

Abstract

Pose estimation is a vital step in many robotics and perception tasks such as robotic manipulation, autonomous vehicle navigation, etc. Current state-of-the-art pose estimation methods rely on deep neural networks with complicated structures and long inference times. While highly robust, they require computing power often unavailable on mobile robots. We propose a CNN-based pose refinement system which takes a coarsely estimated 3D pose from a computationally cheaper algorithm along with a bounding box image of the object, and returns a highly refined pose. Our experiments on the YCB-Video dataset show that our system can refine 3D poses to an extremely high precision with minimal training data.