ViewpointDepth: A New Dataset for Monocular Depth Estimation Under Viewpoint Shifts
This work addresses a critical issue for autonomous driving and computer vision by providing a dataset to assess model robustness to geometric shifts, though it is incremental as it focuses on evaluation rather than a new method.
The paper tackles the underexplored problem of viewpoint shifts affecting monocular depth estimation models by introducing a new dataset and evaluation methodology, revealing limitations in current models and emphasizing the importance of viewpoint variations in real-world applications.
Monocular depth estimation is a critical task for autonomous driving and many other computer vision applications. While significant progress has been made in this field, the effects of viewpoint shifts on depth estimation models remain largely underexplored. This paper introduces a novel dataset and evaluation methodology to quantify the impact of different camera positions and orientations on monocular depth estimation performance. We propose a ground truth strategy based on homography estimation and object detection, eliminating the need for expensive LIDAR sensors. We collect a diverse dataset of road scenes from multiple viewpoints and use it to assess the robustness of a modern depth estimation model to geometric shifts. After assessing the validity of our strategy on a public dataset, we provide valuable insights into the limitations of current models and highlight the importance of considering viewpoint variations in real-world applications.