Text-Visual Prompting for Efficient 2D Temporal Video Grounding | Read Paper on Bytez