Invert4TVG: A Temporal Video Grounding Framework with Inversion Tasks for Enhanced Action Understanding

Open in new window