Visual Grounding on RefCOCO (TestB)
92.6AccuracyVanilla
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| VanillaBackbone Model=DeepSeek-VL2 Small-16B, Average Tokens=100%2026.02 | 92.6 | — | — | — | — | — | — | — | |
| InternVL2.5Number of Parameters=78B2025.02 | 92.5 | — | — | — | — | — | — | — | |
| Florence-2Backbone=DaViT-B2026.01 | 92 | — | — | — | — | — | — | — | |
| Florence-2-L#PT=126M, #FT=127K2025.03 | 92 | — | — | — | — | — | — | — | |
| IVC-PruneBackbone Model=DeepSeek-VL2 Small-16B, Average Tokens=52%2026.02 | 91.8 | — | — | — | — | — | — | — | |
| UNINEXT#PT=600K, #FT=127K2025.03 | 91.5 | — | — | — | — | — | — | — | |
| Florence-2-B#PT=126M, #FT=127K2025.03 | 91.5 | — | — | — | — | — | — | — | |
| UNINEXTModel Category=Specialist2024.03 | 91.46 | — | — | — | — | — | — | — | |
| Qwen3-VL-8B-Thinking + EGMBackbone=Qwen3-VL-8B, Enhancement=EGM2026.01 | 91.2 | — | — | — | — | — | — | — | |
| Qwen3-VL-4B-Thinking + EGMBackbone=Qwen3-VL-4B, Enhancement=EGM2026.01 | 90.9 | — | — | — | — | — | — | — | |
| Youtu-VLModel Size=4B2026.01 | 90.8 | — | — | — | — | — | — | — | |
| Qwen3-VL-235B-A22B-ThinkingBackbone=Qwen3-VL-235B-A22B, Variant=Thinking2026.01 | 90.6 | — | — | — | — | — | — | — | |
| W-FastVBackbone Model=DeepSeek-VL2 Small-16B, Average Tokens=54%2026.02 | 90.4 | — | — | — | — | — | — | — | |
| InternVL-3-78BBackbone=InternVL-3-78B2026.01 | 90.3 | — | — | — | — | — | — | — | |
| Qwen3-VL-235B-A22B-InstructBackbone=Qwen3-VL-235B-A22B, Variant=Instruct2026.01 | 90.2 | — | — | — | — | — | — | — | |
| InternVL-3-8B + EGMBackbone=InternVL-3-8B, Enhancement=EGM2026.01 | 90.1 | — | — | — | — | — | — | — | |
| Qwen2.5-VLNumber of Parameters=72B2025.02 | 89.7 | — | — | — | — | — | — | — | |
| FastVBackbone Model=DeepSeek-VL2 Small-16B, Average Tokens=54%2026.02 | 89.5 | — | — | — | — | — | — | — | |
| Qwen3-VL-2B-Thinking + EGMBackbone=Qwen3-VL-2B, Enhancement=EGM2026.01 | 89.4 | — | — | — | — | — | — | — | |
| InfMLLM-7BLLM=Vicuna-7B2023.11 | 89.3 | — | — | — | — | — | — | — | |
| ONE-PEACE2023.05 | 89.26 | — | — | — | — | — | — | — | |
| UFOBackbone=InternVL2.5-8B2026.01 | 89.2 | — | — | — | — | — | — | — | |
| PDropBackbone Model=DeepSeek-VL2 Small-16B, Average Tokens=57%2026.02 | 89.1 | — | — | — | — | — | — | — | |
| CogVLM-Grounding-17BModel Size=17B, Zero-shot=false2024.02 | 88.99 | — | — | — | — | — | — | — | |
| Ferretv2#PT=1.1M, #FT=127K2025.03 | 88.7 | — | — | — | — | — | — | — | |
| OFA2023.05 | 88.44 | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B AIFAIF=true2026.04 | 88.4 | — | — | — | — | — | — | — | |
| Grounding-DINO2023.05 | 88.24 | — | — | — | — | — | — | — | |
| G-DINO-LModel Category=Specialist2024.03 | 88.24 | — | — | — | — | — | — | — | |
| Grounding DINO2025.02 | 88.2 | — | — | — | — | — | — | — | |
| Grounding DINOBackbone=Swin-L2026.01 | 88.2 | — | — | — | — | — | — | — | |
| InternVL-3.5Model Size=4B2026.01 | 88.2 | — | — | — | — | — | — | — | |
| Grounding-DINO-L2026.04 | 88.2 | — | — | — | — | — | — | — | |
| InternVL-3-8BBackbone=InternVL-3-8B2026.01 | 88 | — | — | — | — | — | — | — | |
| Qwen3-VL-8B-InstructBackbone=Qwen3-VL-8B, Variant=Instruct2026.01 | 87.8 | — | — | — | — | — | — | — | |
| Qwen3-VL-4B-InstructBackbone=Qwen3-VL-4B, Variant=Instruct2026.01 | 87.6 | — | — | — | — | — | — | — | |
| InternVL-3-2B + EGMBackbone=InternVL-3-2B, Enhancement=EGM2026.01 | 87.4 | — | — | — | — | — | — | — | |
| Phi3V-dist#PT=0, #FT=127K2025.03 | 87.3 | — | — | — | — | — | — | — | |
| FIBER2023.05 | 87.26 | — | — | — | — | — | — | — | |
| InternVL-3-1B + EGMBackbone=InternVL-3-1B, Enhancement=EGM2026.01 | 87 | — | — | — | — | — | — | — | |
| Qwen3-VL-8B-ThinkingBackbone=Qwen3-VL-8B, Variant=Thinking2026.01 | 86.9 | — | — | — | — | — | — | — | |
| Qwen3-VLModel Size=4B2026.01 | 86.7 | — | — | — | — | — | — | — | |
| InternVL-3-2BBackbone=InternVL-3-2B2026.01 | 86.4 | — | — | — | — | — | — | — | |
| GriffonBackbone=LLama2-13B2026.01 | 86.1 | — | — | — | — | — | — | — | |
| VanillaBackbone Model=InternVL 2.5 8B, Average Tokens=100%2026.02 | 86 | — | — | — | — | — | — | — | |
| Phi3V-sft#PT=0, #FT=127K2025.03 | 86 | — | — | — | — | — | — | — | |
| Phi3V-vocab#PT=0, #FT=127K2025.03 | 86 | — | — | — | — | — | — | — | |
| InternVL-2.5-8B2026.04 | 85.9 | — | — | — | — | — | — | — | |
| IVC-PruneBackbone Model=InternVL 2.5 8B, Average Tokens=50%2026.02 | 85.7 | — | — | — | — | — | — | — | |
| Qwen3-VL-4B-ThinkingBackbone=Qwen3-VL-4B, Variant=Thinking2026.01 | 85.6 | — | — | — | — | — | — | — | |
| Qwen2.5-VLNumber of Parameters=7B2025.02 | 85.4 | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B2026.04 | 85.4 | — | — | — | — | — | — | — | |
| QWen-VL-7BLLM=Qwen-7B2023.11 | 85.34 | — | — | — | — | — | — | — | |
| Qwen-VL-7BModel Size=7B, Zero-shot=false2024.02 | 85.34 | — | — | — | — | — | — | — | |
| MiniGPT-v2-7BLLM=LLaMA2-7B2023.11 | 85.33 | — | — | — | — | — | — | — | |
| VanillaBackbone=QWEN-2.5-VL 7B, Average Tokens=100%2026.02 | 85.16 | — | — | — | — | — | — | — | |
| VanillaBackbone Model=Qwen2.5-VL 7B, Average Tokens=100%2026.02 | 84.7 | — | — | — | — | — | — | — | |
| Qwen3-VL-2B-InstructBackbone=Qwen3-VL-2B, Variant=Instruct2026.01 | 84.7 | — | — | — | — | — | — | — | |
| IVC-PruneBackbone Model=Qwen2.5-VL 7B, Average Tokens=50%2026.02 | 84.5 | — | — | — | — | — | — | — | |
| Ferret#PT=1.1M, #FT=127K2025.03 | 84.4 | — | — | — | — | — | — | — | |
| NuwaBackbone=QWEN-2.5-VL 7B, Average Tokens=75%2026.02 | 84.37 | — | — | — | — | — | — | — | |
| VisionLLM v2Backbone=Swin-T2026.01 | 84.3 | — | — | — | — | — | — | — | |
| Qwen2.5-VLNumber of Parameters=3B2025.02 | 84 | — | — | — | — | — | — | — | |
| SeqTRVisual Encoder=DN53, Params (M)=7.90, Pre-train images=174K2022.03 | 83.59 | — | — | — | — | — | — | — | |
| Lumen-7BModel Category=Generalist2024.03 | 83.55 | — | — | — | — | — | — | — | |
| Localization Heads2026.04 | 83.3 | — | — | — | — | — | — | — | |
| UNICORN2023.05 | 83.06 | — | — | — | — | — | — | — | |
| VistaLLM-7BModel Size=7B, Zero-shot=false2024.02 | 83 | — | — | — | — | — | — | — | |
| NuwaBackbone=QWEN-2.5-VL 7B, Average Tokens=50%2026.02 | 82.85 | — | — | — | — | — | — | — | |
| MDETRBackbone=ENB32026.01 | 82.7 | — | — | — | — | — | — | — | |
| Ferret-7BModel Size=7B, Zero-shot=false2024.02 | 82.45 | — | — | — | — | — | — | — | |
| CoLLaVO-7BModel Size=7B, Zero-shot=true2024.02 | 82.39 | — | — | — | — | — | — | — | |
| QRNet2024.01 | 82.34 | — | — | — | — | — | — | — | |
| VScanBackbone Model=Qwen2.5-VL 7B, Average Tokens=50%2026.02 | 82.2 | — | — | — | — | — | — | — | |
| Qwen3-VL-2B-ThinkingBackbone=Qwen3-VL-2B, Variant=Thinking2026.01 | 82 | — | — | — | — | — | — | — | |
| Shikra-13BLLM=Vicuna-13B2023.11 | 81.81 | — | — | — | — | — | — | — | |
| Shikra2026.02 | 81.8 | — | — | — | — | — | — | — | |
| UVLTrackModel Version=B2024.01 | 81.73 | — | — | — | — | — | — | — | |
| InternVL-3-1BBackbone=InternVL-3-1B2026.01 | 81.7 | — | — | — | — | — | — | — | |
| MDETRVisual Encoder=RN101, Params (M)=17.36, Pre-train images=200K2022.03 | 81.41 | — | — | — | — | — | — | — | |
| MDETR2023.05 | 81.41 | — | — | — | — | — | — | — | |
| SeqTR2024.01 | 81.24 | — | — | — | — | — | — | — | |
| Fully-FTArchitecture=MDETR2023.08 | 81.22 | — | — | — | — | — | — | — | |
| Fully-FTParams.=185.2, Mem.=39.62026.04 | 81.22 | — | — | — | — | — | — | — | |
| Fully-FTParams. (M)=185.2, Mem. (G)=39.62026.04 | 81.22 | — | — | — | — | — | — | — | |
| RefTRVisual Encoder=RN101, Params (M)=17.86, Pre-train images=100K2022.03 | 81.16 | — | — | — | — | — | — | — | |
| APE (D)Backbone=ViT-L2023.12 | 80.9 | 68.4 | — | — | — | — | — | — | |
| InternVL2-8B2026.04 | 80.7 | — | — | — | — | — | — | — | |
| VLTVG2024.01 | 80.49 | — | — | — | — | — | — | — | |
| Shikra-7BLLM=Vicuna-7B2023.11 | 80.24 | — | — | — | — | — | — | — | |
| Shikra-7BModel Category=Generalist2024.03 | 80.24 | — | — | — | — | — | — | — | |
| Shikra-7BModel Size=7B, Zero-shot=false2024.02 | 80.24 | — | — | — | — | — | — | — | |
| Shikra-7B2026.04 | 80.2 | — | — | — | — | — | — | — | |
| W-FastVBackbone Model=Qwen2.5-VL 7B, Average Tokens=54%2026.02 | 79.8 | — | — | — | — | — | — | — | |
| Partially⇑Architecture=MDETR2023.08 | 79.78 | — | — | — | — | — | — | — | |
| Ours†Params. (M)=0.8, Mem. (G)=7.32026.04 | 79.2 | — | — | — | — | — | — | — | |
| Ours‡Params. (M)=0.8, Mem. (G)=7.42026.04 | 79.19 | — | — | — | — | — | — | — | |
| Ours♠Params.=0.8, Mem.=7.1, Time (ms)=52.292026.04 | 78.97 | — | — | — | — | — | — | — | |
| Ours♡Params.=0.7, Mem.=7.1, Time (ms)=52.632026.04 | 78.69 | — | — | — | — | — | — | — | |
| SHERLParams.=0.7, Mem.=6.8, Time (ms)=79.332026.04 | 78.41 | — | — | — | — | — | — | — |