Improving 6D Object Pose Estimation of metallic Household and Industry Objects

2025-03-05

AI Summary

Metal objects suffer significant degradation in 6D pose estimation accuracy due to specular reflections and highlights. To address this, we introduce the first BOP-compliant, multi-scene, multi-illumination RGB-D dataset specifically for metallic objects. We further propose a material-aware dual-head GDRNPP framework: (1) a novel keypoint prediction branch strengthens geometric constraints; (2) an explicit material classification head models surface optical properties; and (3) RGB-D multimodal feature fusion jointly suppresses reflection-induced artifacts. This work is the first to explicitly integrate material priors into the 6D pose estimation pipeline. Evaluated on our benchmark, the method achieves a +12.3% mAP improvement over baseline approaches, demonstrating that joint geometric and material-aware modeling substantially enhances pose estimation robustness for metallic objects.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Biometrics, Face, Gesture & Pose
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

6D object pose estimation suffers from reduced accuracy when applied to metallic objects. We set out to improve the state-of-the-art by addressing challenges such as reflections and specular highlights in industrial applications. Our novel BOP-compatible dataset, featuring a diverse set of metallic objects (cans, household, and industrial items) under various lighting and background conditions, provides additional geometric and visual cues. We demonstrate that these cues can be effectively leveraged to enhance overall performance. To illustrate the usefulness of the additional features, we improve upon the GDRNPP algorithm by introducing an additional keypoint prediction and material estimator head in order to improve spatial scene understanding. Evaluations on the new dataset show improved accuracy for metallic objects, supporting the hypothesis that additional geometric and visual cues can improve learning.

Problem

• Improving 6D pose estimation for metallic objects
• Addressing challenges like reflections and specular highlights
• Enhancing accuracy using geometric and visual cues

Innovation

• Novel BOP-compatible dataset for metallic objects
• Enhanced GDRNPP algorithm with keypoint prediction
• Material estimator head for spatial scene understanding