STORM: Segment, Track, and Object Re-Localization from a Single Image
STORM is a unified framework for 6D pose tracking using only a single reference image, eliminating manual masking. It combines Hierarchical Spatial Fusion Attention to align query images with references and a BCE-trained Tracking Object Module to detect tracking drift. This allows the system to reliably recover from occlusions and fast motion by triggering automatic re-initialization.