List your company
ResearchFormerly WizWand

Reasoning Segmentation on EarthReason (test)

76.86gIoUGRTO
51.556858.125964.69571.2641Dec 22, 2025Jan 16, 2026Feb 10, 2026Mar 7, 2026Apr 1, 2026Apr 26, 2026May 22, 2026
Updated 4mo ago

Evaluation results

MethodLinks
GRTO2026.0576.8679.3——
B-GRTO2026.0576.2778.54——
GRPO2026.0574.3176.06——
Think2Seg-RSModel=Qwen-2.5-VL-7B + SAM2-Small2025.1273.7375.83——
SegEarth-R2-3BLLM=Phi-2-2.7B2025.1273.569.5——
Think2Seg-RS2026.0573.3675.6——
UniGeoSegdomain-specific=true2026.0573.0874.61——
Text4Seg++LLM=Qwen2-7B2025.127365.6——
Think2Seg-RSModel=Qwen-2.5-VL-3B + SAM2-Small2025.1271.0974.13——
RemoteReasonerLLM=Qwen2.5-7B2025.127169.1——
RemoteReasonerModel=Qwen-2.5-VL-7B + SAM22025.1270.9669.13——
RemoteReasonerdomain-specific=true2026.0570.9669.02——
SegEarth-R1LLM=Phi-1.5-1.3B2025.1270.868.3——
SegEarth-R1Model=Swin-B (88M) + Phi-1.5-1.3B2025.1270.7568.25——
SegEarth-R1domain-specific=true2026.0570.7568.6——
PSALMModel=Swin-B (88M) + Phi-1.5-1.3B2025.1268.364.61——
PSALM2026.0568.364.61——
Seg-Zero2026.0567.8962.43——
VisionReasoner2026.0567.1965.41——
VisionReasonerModel=Qwen-2.5-VL-7B + SAM2-Large2025.1267.1765.29——
Seg-ZeroModel=Qwen-2.5-VL-7B + SAM2-Large2025.1263.1662.76——
LISALLM=Vicuna-7B2025.1260.959.1——
LISAModel=CLIP-L (304M) + Vicuna-7B2025.1260.8859.1——
LISA2026.0560.8861.04——
PixelLMModel=CLIP-L (304M) + Vicuna-7B2025.1260.0159.22——
PixelLM2026.0560.0159.22——
PixelLMLLM=Vicuna-7B2025.126059.2——
GeoPixeldomain-specific=true2026.0552.5353.9——
Earth-OneVisionScale=2B2026.06——57.9763.41
Earth-OneVisionScale=2B, Auxiliary=SAM22026.06——74.262.34
GeoPixelScale=7B2026.06——52.5353.9
GPT-5-SAMLLM=GPT-5, Trained on RS data: LLM=false, Trained on RS data: Mask Decoder=false, Trained on RS data: Extra=false2026.02——47.4—
GPT-Image-1LLM=GPT-5, Trained on RS data: LLM=false, Trained on RS data: Mask Decoder=false, Trained on RS data: Extra=false2026.02——41—
LISALLM=Vicuna-7B, Trained on RS data: LLM=LoRA, Trained on RS data: Mask Decoder=true, Trained on RS data: Extra=true2026.02——60.9—
LISAScale=7B2026.06——60.8861.04
PixelLMLLM=Vicuna-7B, Trained on RS data: LLM=LoRA, Trained on RS data: Mask Decoder=true, Trained on RS data: Extra=true2026.02——60—
PixelLMScale=13B2026.06——60.0157.94
PSALMLLM=phi-1.5-1.3B, Trained on RS data: LLM=true, Trained on RS data: Mask Decoder=true, Trained on RS data: Extra=true2026.02——68.3—
PSALMScale=1.3B2026.06——68.366.61
Qwen3-VL-SAMLLM=Qwen3-VL-2B, Trained on RS data: LLM=LoRA, Trained on RS data: Mask Decoder=false, Trained on RS data: Extra=false2026.02——72.7—
RemoteReasonerScale=7B, Auxiliary=SAM22026.06——70.9669.02
SegEarth-R1LLM=phi-1.5-1.3B, Trained on RS data: LLM=true, Trained on RS data: Mask Decoder=true, Trained on RS data: Extra=true2026.02——70.7—
SegEarth-R1Scale=1.3B2026.06——70.7568.6