Implicit 3D Orientation Learning for 6D Object Detection from RGB Images

TL;DR

This paper introduces a real-time RGB-based 6D object detection pipeline using a novel Augmented Autoencoder for implicit 3D orientation learning, eliminating the need for pose-annotated training data and handling symmetries effectively.

Contribution

It presents a new autoencoder-based method for 3D orientation estimation that generalizes across sensors and symmetries without requiring real pose annotations.

Findings

Achieves state-of-the-art results on T-LESS dataset.

Performs competitively on LineMOD dataset.

Handles object and view symmetries inherently.

Abstract

We propose a real-time RGB-based pipeline for object detection and 6D pose estimation. Our novel 3D orientation estimation is based on a variant of the Denoising Autoencoder that is trained on simulated views of a 3D model using Domain Randomization. This so-called Augmented Autoencoder has several advantages over existing methods: It does not require real, pose-annotated training data, generalizes to various test sensors and inherently handles object and view symmetries. Instead of learning an explicit mapping from input images to object poses, it provides an implicit representation of object orientations defined by samples in a latent space. Our pipeline achieves state-of-the-art performance on the T-LESS dataset both in the RGB and RGB-D domain. We also evaluate on the LineMOD dataset where we can compete with other synthetically trained approaches. We further increase performance by correcting 3D orientation estimates to account for perspective errors when the object deviates from the image center and show extended results.