Deep-6DPose: Recovering 6D Object Pose from a Single RGB Image

Abstract

Detecting objects and their 6D poses from only RGB images is an important task for many robotic applications. While deep learning methods have made significant progress in visual object detection and segmentation, the object pose estimation task is still challenging. In this paper, we introduce an end-toend deep learning framework, named Deep-6DPose, that jointly detects, segments, and most importantly recovers 6D poses of object instances from a single RGB image. In particular, we extend the recent state-of-the-art instance segmentation network Mask R-CNN with a novel pose estimation branch to directly regress 6D object poses without any post-refinements. Our key technical contribution is the decoupling of pose parameters into translation and rotation so that the rotation can be regressed via a Lie algebra representation. The resulting pose regression loss is differential and unconstrained, making the training tractable. The experiments on two standard pose benchmarking datasets show that our proposed approach compares favorably with the state-of-the-art RGB-based multi-stage pose estimation methods. Importantly, due to the end-to-end architecture, Deep-6DPose is considerably faster than competing multi-stage methods, offers an inference speed of 10 fps that is well suited for robotic applications.

References in corpus (3)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Learning Analysis-by-Synthesis for 6D Pose Estimation in RGB-D Images

Cited by in corpus (18)

• On the Continuity of Rotation Representations in Neural Networks
• REDE: End-to-end Object 6D Pose Robust Estimation Using Differentiable Outliers Elimination
• Deep Learning for Spacecraft Pose Estimation from Photorealistic Rendering
• Multi-Task Deep Networks for Depth-Based 6D Object Pose and Joint Registration in Crowd Scenarios
• Learning Canonical Shape Space for Category-Level 6D Object Pose and Size Estimation
• Semantic keypoint-based pose estimation from single RGB frames
• 6D Object Pose Estimation without PnP
• Pose Proposal Critic: Robust Pose Refinement by Learning Reprojection Errors
• ConvPoseCNN2: Prediction and Refinement of Dense 6D Object Poses
• Learning Neural Representation of Camera Pose with Matrix Representation of Pose Shift via View Synthesis
• PrimA6D: Rotational Primitive Reconstruction for Enhanced and Robust 6D Pose Estimation
• Vehicle Reconstruction and Texture Estimation Using Deep Implicit Semantic Template Mapping
• StablePose: Learning 6D Object Poses from Geometrically Stable Patches
• Occlusion Resistant Object Rotation Regression from Point Cloud Segments
• Projective Manifold Gradient Layer for Deep Rotation Regression
• 3D Point-to-Keypoint Voting Network for 6D Pose Estimation
• Orienting Novel 3D Objects Using Self-Supervised Learning of Rotation Transforms
• L6DNet: Light 6 DoF Network for Robust and Precise Object Pose Estimation with Small Datasets