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

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.

PDF Abstract ECCV 2018 PDF ECCV 2018 Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
6D Pose Estimation using RGB LineMOD Augmented Autoencoder Mean ADD 28.7 # 21
6D Pose Estimation using RGBD LineMOD Augmented Autoencoder Mean ADD 64.67 # 7
6D Pose Estimation using RGBD T-LESS Augmented Autoencoder Mean Recall 72.76 # 1
6D Pose Estimation using RGB T-LESS Augmented Autoencoder Mean Recall 36.8 # 1

Methods