OnePose: One-Shot Object Pose Estimation without CAD Models

TL;DR

OnePose introduces a novel object pose estimation method that does not require CAD models, using a visual localization-inspired approach with a graph attention network for real-time, category-agnostic 6D pose tracking from RGB videos.

Contribution

It presents a CAD-model-free, category-agnostic pose estimation framework utilizing a graph attention network for efficient 6D pose tracking from RGB videos.

Findings

Achieves real-time 6D pose detection and tracking.

Handles arbitrary object categories without CAD models.

Introduces a large-scale dataset with 150 objects and 450 sequences.

Abstract

We propose a new method named OnePose for object pose estimation. Unlike existing instance-level or category-level methods, OnePose does not rely on CAD models and can handle objects in arbitrary categories without instance- or category-specific network training. OnePose draws the idea from visual localization and only requires a simple RGB video scan of the object to build a sparse SfM model of the object. Then, this model is registered to new query images with a generic feature matching network. To mitigate the slow runtime of existing visual localization methods, we propose a new graph attention network that directly matches 2D interest points in the query image with the 3D points in the SfM model, resulting in efficient and robust pose estimation. Combined with a feature-based pose tracker, OnePose is able to stably detect and track 6D poses of everyday household objects in real-time. We also collected a large-scale dataset that consists of 450 sequences of 150 objects.