Deep Weakly Supervised Positioning
arXiv:2104.04866
Abstract
PoseNet can map a photo to the position where it is taken, which is appealing in robotics. However, training PoseNet requires full supervision, where ground truth positions are non-trivial to obtain. Can we train PoseNet without knowing the ground truth positions for each observation? We show that this is possible via constraint-based weak-supervision, leading to the proposed framework: DeepGPS. Particularly, using wheel-encoder-estimated distances traveled by a robot along random straight line segments as constraints between PoseNet outputs, DeepGPS can achieve a relative positioning error of less than 2%. Moreover, training DeepGPS can be done as auto-calibration with almost no human attendance, which is more attractive than its competing methods that typically require careful and expert-level manual calibration. We conduct various experiments on simulated and real datasets to demonstrate the general applicability, effectiveness, and accuracy of DeepGPS, and perform a comprehensive analysis of its robustness. Our code is available at https://ai4ce.github.io/DeepGPS/.
arXiv comment8 pages, 8 figures, submitted to IEEE Robotics and Automation Letters (RA-L) and 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021)
References in corpus (11)
- Adam: A Method for Stochastic Optimization
- Going Deeper with Convolutions
- DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks
- Unsupervised Learning of Depth and Ego-Motion from Video
- InLoc: Indoor Visual Localization with Dense Matching and View Synthesis
- VidLoc: A Deep Spatio-Temporal Model for 6-DoF Video-Clip Relocalization
- Understanding the Limitations of CNN-based Absolute Camera Pose Regression
- DeepMapping: Unsupervised Map Estimation From Multiple Point Clouds
- EMPNet: Neural Localisation and Mapping Using Embedded Memory Points
- Geometry-Aware Learning of Maps for Camera Localization
- Learning by Inertia: Self-supervised Monocular Visual Odometry for Road Vehicles