SpatialVID: A Large-Scale Video Dataset with Spatial Annotations
This dataset addresses a key bottleneck for researchers in video and 3D vision by providing scalable, diverse data to improve model generalization and performance.
The authors tackled the problem of limited large-scale, high-quality training data for spatial intelligence in videos by introducing SpatialVID, a dataset with over 21,000 hours of raw video processed into 2.7 million clips totaling 7,089 hours, enriched with dense 3D annotations like camera poses and depth maps.
Significant progress has been made in spatial intelligence, spanning both spatial reconstruction and world exploration. However, the scalability and real-world fidelity of current models remain severely constrained by the scarcity of large-scale, high-quality training data. While several datasets provide camera pose information, they are typically limited in scale, diversity, and annotation richness, particularly for real-world dynamic scenes with ground-truth camera motion. To this end, we collect \textbf{SpatialVID}, a dataset consists of a large corpus of in-the-wild videos with diverse scenes, camera movements and dense 3D annotations such as per-frame camera poses, depth, and motion instructions. Specifically, we collect more than 21,000 hours of raw video, and process them into 2.7 million clips through a hierarchical filtering pipeline, totaling 7,089 hours of dynamic content. A subsequent annotation pipeline enriches these clips with detailed spatial and semantic information, including camera poses, depth maps, dynamic masks, structured captions, and serialized motion instructions. Analysis of SpatialVID's data statistics reveals a richness and diversity that directly foster improved model generalization and performance, establishing it as a key asset for the video and 3D vision research community.