CamoSAM2: SAM2-oriented Prompt Auto-Refinement for Video Camouflaged Object Detection
Researchers have developed new frameworks for camouflaged object detection (COD) that address the issue of over-detection. One approach, CFCamo, uses a counterfactual benchmark to train agents to both detect camouflaged objects and abstain when no object is present, improving performance on existing datasets and achieving high pair accuracy on the new CF-COD benchmark. Another method, CamoSAM2, refines prompts for the Segment Anything Model 2 (SAM2) by integrating motion and appearance cues to enhance automatic detection and segmentation of camouflaged objects in videos, outperforming current state-of-the-art methods in mean intersection over union (mIoU) and inference speed. AI
IMPACT These advancements in camouflaged object detection could improve AI's ability to accurately identify and segment objects in complex visual environments, impacting fields like surveillance, medical imaging, and autonomous systems.