Unsupervised Metric Relocalization Using Transform Consistency Loss
arXiv:2011.00608
Abstract
Training networks to perform metric relocalization traditionally requires accurate image correspondences. In practice, these are obtained by restricting domain coverage, employing additional sensors, or capturing large multi-view datasets. We instead propose a self-supervised solution, which exploits a key insight: localizing a query image within a map should yield the same absolute pose, regardless of the reference image used for registration. Guided by this intuition, we derive a novel transform consistency loss. Using this loss function, we train a deep neural network to infer dense feature and saliency maps to perform robust metric relocalization in dynamic environments. We evaluate our framework on synthetic and real-world data, showing our approach outperforms other supervised methods when a limited amount of ground-truth information is available.
arXiv commentAccepted for publication in the 4th Conference on Robot Learning (CoRL 2020), Cambridge MA, USA
References in corpus (9)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimator
- The Cityscapes Dataset for Semantic Urban Scene Understanding
- TILDE: A Temporally Invariant Learned DEtector
- CARLA: An Open Urban Driving Simulator
- Universal Correspondence Network
- Unsupervised Learning of Depth and Ego-Motion from Video
- LDSO: Direct Sparse Odometry with Loop Closure
- Taking a Deeper Look at the Inverse Compositional Algorithm