The SAM2-to-SAM3 Gap in the Segment Anything Model Family: Why Prompt-Based Expertise Fails in Concept-Driven Image Segmentation

This paper investigates the fundamental discontinuity between the latest two Segment Anything Models: SAM2 and SAM3 (also called SAMv2 and SAMv3). We explain why the expertise in prompt-based segmentation of SAM2 does not transfer to the multimodal concept-driven paradigm of SAMv3. SAMv2 operates through spatial prompts points, boxes, and masks yielding purely geometric and temporal segmentation. In contrast, SAMv3 introduces a unified vision-language architecture capable of open-vocabulary reasoning, semantic grounding, contrastive alignment, and exemplar-based concept understanding. We structure this analysis through five core components: (1) a Conceptual Break Between Prompt-Based and Concept-Based Segmentation, contrasting spatial prompt semantics of SAMv2 with multimodal fusion and text-conditioned mask generation of SAMv3; (2) Architectural Divergence, detailing pure vision-temporal design of SAMv2 versus integration of vision-language encoders, geometry and exemplar encoders, fusion modules, DETR-style decoders, object queries, and ambiguity-handling via Mixture-of-Experts in SAMv3; (3) Dataset and Annotation Differences, contrasting SA-V video masks with multimodal concept-annotated corpora of SAMv3; (4) Training and Hyperparameter Distinctions, showing why SAMv2 optimization knowledge does not apply to SAM3; and (5) Evaluation, Metrics, and Failure Modes, outlining the transition from geometric IoU metrics to semantic, open-vocabulary evaluation. Together, these analyses establish SAMv3 as a new class of segmentation foundation model and chart future directions for the emerging concept-driven segmentation era.