SOTA Reasoning Segmentation benchmarks and papers with code

Benchmarks

Dataset Name	SOTA Method	Metric
ReasonSeg (val)	Vision Banana	gIoU79.3	327	1d ago
ReasonSeg (test)	TikArt	gIoU73.8	236	1d ago
EarthReason (val)	GRTO	gIoU76.35	47	9d ago
ReasonSeg long (test)	EVOL-SAM3	cIoU0.733	40	3mo ago
ReasonSeg short (test)	SAM 3 Agent	cIoU59.4	40	3mo ago
ReasonSeg	InstructSeg	gIoU61.9	33	1mo ago
EarthReason (test)	GRTO	gIoU76.86	28	9d ago
gRefCOCO (testB)	SegCompass-13B	gIoU0.719	22	12d ago
gRefCOCO (testA)	SegCompass-13B	gIoU77.7	22	12d ago
MUSE (val)	POPEN	gIoU (overall)48	21	3mo ago
MUSE (test)	PixelLM-Llama2-13B	gIoU (overall)42.3	16	3mo ago
SegEarth-R2 (train)	LISA-7B	Mean IoU26	14	3mo ago
GeoSeg-Bench	GeoSeg	IoU56.4	14	3mo ago
ReasonSeg-X (test)	StAR	gIoU57.9	12	9d ago
ReasonSeg-X (val)	StAR	gIoU62.6	12	9d ago
ReasonSeg-R (test)	StAR	gIoU74.9	12	9d ago
DRSeg	PixDLM	Attribute Reasoning gIoU62.8	12	1mo ago
ReasonSeg zero-shot (val)	WISE-7B	gIoU63.5	11	2mo ago
ReasonSeg zero-shot (test)	WISE-7B	gIoU60.3	10	2mo ago
Intent2Part InstructPart	SegWorld v2	mIoU77.2	9	6d ago
HALLUSEGBENCH Reasoning	VisionReasoner	CMS Factual0.7315	9	1mo ago
MVOR Reason	ORDIRS	cIoU (Semantic)83.21	8	3mo ago
In-house OR video dataset 1.0 (test)	ORDIRS-Agent	cIoU (Semantic)75.8	8	3mo ago
ReasonSeg (combined)	VisHarness	Average Score68.6	6	5d ago
REASONSEG HARD (test)	CoT-Seg	gIoU58.6	6	3mo ago

Showing 25 of 26 rows