Spatio-Temporal Video Grounding Tasks | State-of-the-Art