Add detector Segment Anything Model (SAMv2) to facilitate text-driven object detection and segmentation. Fore example to automatically generate segmentation masks for smoke by inputting the prompt ”smoke.”