MobilePose: Real-Time Pose Estimation for Unseen Objects with Weak Shape Supervision
arXiv:2003.03522
Abstract
In this paper, we address the problem of detecting unseen objects from RGB images and estimating their poses in 3D. We propose two mobile friendly networks: MobilePose-Base and MobilePose-Shape. The former is used when there is only pose supervision, and the latter is for the case when shape supervision is available, even a weak one. We revisit shape features used in previous methods, including segmentation and coordinate map. We explain when and why pixel-level shape supervision can improve pose estimation. Consequently, we add shape prediction as an intermediate layer in the MobilePose-Shape, and let the network learn pose from shape. Our models are trained on mixed real and synthetic data, with weak and noisy shape supervision. They are ultra lightweight that can run in real-time on modern mobile devices (e.g. 36 FPS on Galaxy S20). Comparing with previous single-shot solutions, our method has higher accuracy, while using a significantly smaller model (2~3% in model size or number of parameters).
References in corpus (7)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- YOLO9000: Better, Faster, Stronger
- Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects
- MediaPipe: A Framework for Building Perception Pipelines
- HybridPose: 6D Object Pose Estimation under Hybrid Representations
- Object as Distribution
Cited by in corpus (8)
- A Survey on Deep Learning-Based Monocular Spacecraft Pose Estimation: Current State, Limitations and Prospects
- Optimal and Robust Category-level Perception: Object Pose and Shape Estimation from 2D and 3D Semantic Keypoints
- Real-Time, Flight-Ready, Non-Cooperative Spacecraft Pose Estimation Using Monocular Imagery
- Optimal Pose and Shape Estimation for Category-level 3D Object Perception
- SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo
- Instant 3D Object Tracking with Applications in Augmented Reality
- Semantic keypoint-based pose estimation from single RGB frames
- A Flexible-Frame-Rate Vision-Aided Inertial Object Tracking System for Mobile Devices