PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimation

TL;DR

This paper introduces PVN3D, a deep learning approach that detects 3D keypoints from RGBD images to accurately estimate the 6DoF pose of objects, outperforming previous methods on multiple benchmarks.

Contribution

The paper presents a novel deep Hough voting network for 3D keypoint detection and 6DoF pose estimation from RGBD images, extending 2D keypoint methods to 3D with improved accuracy.

Findings

Outperforms state-of-the-art methods on several benchmarks.

Effectively utilizes geometric constraints of rigid objects.

Demonstrates robustness in 6DoF pose estimation from RGBD data.

Abstract

In this work, we present a novel data-driven method for robust 6DoF object pose estimation from a single RGBD image. Unlike previous methods that directly regressing pose parameters, we tackle this challenging task with a keypoint-based approach. Specifically, we propose a deep Hough voting network to detect 3D keypoints of objects and then estimate the 6D pose parameters within a least-squares fitting manner. Our method is a natural extension of 2D-keypoint approaches that successfully work on RGB based 6DoF estimation. It allows us to fully utilize the geometric constraint of rigid objects with the extra depth information and is easy for a network to learn and optimize. Extensive experiments were conducted to demonstrate the effectiveness of 3D-keypoint detection in the 6D pose estimation task. Experimental results also show our method outperforms the state-of-the-art methods by large margins on several benchmarks. Code and video are available at https://github.com/ethnhe/PVN3D.git.