360 in the Wild: Dataset for Depth Prediction and View Synthesis
This dataset enables research in depth prediction and view synthesis for panoramic images, addressing a gap for computer vision researchers, but it is incremental as it extends existing dataset types to a new modality.
The authors introduced a large-scale 360-degree video dataset captured from real-world locations to address the lack of panoramic image datasets with pose and depth information, which are mostly synthetic. They demonstrated its relevance for tasks like single image depth estimation and view synthesis, though no specific performance numbers were provided.
The large abundance of perspective camera datasets facilitated the emergence of novel learning-based strategies for various tasks, such as camera localization, single image depth estimation, or view synthesis. However, panoramic or omnidirectional image datasets, including essential information, such as pose and depth, are mostly made with synthetic scenes. In this work, we introduce a large scale 360$^{\circ}$ videos dataset in the wild. This dataset has been carefully scraped from the Internet and has been captured from various locations worldwide. Hence, this dataset exhibits very diversified environments (e.g., indoor and outdoor) and contexts (e.g., with and without moving objects). Each of the 25K images constituting our dataset is provided with its respective camera's pose and depth map. We illustrate the relevance of our dataset for two main tasks, namely, single image depth estimation and view synthesis.