Unified Category-Level Object Detection and Pose Estimation from RGB Images using 3D Prototypes

TL;DR

This paper presents a unified neural network framework for category-level object detection and 3D pose estimation from RGB images, achieving state-of-the-art accuracy and robustness without requiring depth data.

Contribution

It introduces the first unified model combining detection and pose estimation for RGB images using neural mesh models and multi-model RANSAC.

Findings

Achieves 22.9% improvement over previous methods on REAL275 dataset.

Demonstrates greater robustness compared to single-stage baselines.

Provides open-source code and models for the community.

Abstract

Recognizing objects in images is a fundamental problem in computer vision. Although detecting objects in 2D images is common, many applications require determining their pose in 3D space. Traditional category-level methods rely on RGB-D inputs, which may not always be available, or employ two-stage approaches that use separate models and representations for detection and pose estimation. For the first time, we introduce a unified model that integrates detection and pose estimation into a single framework for RGB images by leveraging neural mesh models with learned features and multi-model RANSAC. Our approach achieves state-of-the-art results for RGB category-level pose estimation on REAL275, improving on the current state-of-the-art by 22.9% averaged across all scale-agnostic metrics. Finally, we demonstrate that our unified method exhibits greater robustness compared to single-stage baselines. Our code and models are available at https://github.com/Fischer-Tom/unified-detection-and-pose-estimation.