Adversarial Transfer of Pose Estimation Regression
This addresses the generalization problem in camera pose estimation for visual localization applications, representing an incremental improvement through transfer learning techniques.
The paper tackles the problem of camera pose estimation in visual localization, where current regression-based methods show poor generalization across scenes due to dataset shift. The authors develop a deep adaptation network using adversarial learning and self-supervised techniques, demonstrating superiority over baselines and competitive performance with state-of-the-art methods on Cambridge Landmarks and 7Scene datasets.
We address the problem of camera pose estimation in visual localization. Current regression-based methods for pose estimation are trained and evaluated scene-wise. They depend on the coordinate frame of the training dataset and show a low generalization across scenes and datasets. We identify the dataset shift an important barrier to generalization and consider transfer learning as an alternative way towards a better reuse of pose estimation models. We revise domain adaptation techniques for classification and extend them to camera pose estimation, which is a multi-regression task. We develop a deep adaptation network for learning scene-invariant image representations and use adversarial learning to generate such representations for model transfer. We enrich the network with self-supervised learning and use the adaptability theory to validate the existence of scene-invariant representation of images in two given scenes. We evaluate our network on two public datasets, Cambridge Landmarks and 7Scene, demonstrate its superiority over several baselines and compare to the state of the art methods.