List your company
ResearchFormerly WizWand

Referring Expression Comprehension on RefCOCO, RefCOCO+, and RefCOCOg Average

92.3Average AccuracyInternVL2.5-78B
67.932874.258980.58586.9111Apr 19, 2024Aug 15, 2024Dec 12, 2024Apr 10, 2025Aug 6, 2025Dec 3, 2025Apr 1, 2026
Updated 7mo ago

Evaluation results

MethodLinks
InternVL2.5-78BParameters=78B2024.1292.3
Qwen2-VL-72BParameters=72B2024.1291.1
CogVLM-Grounding-17BParameters=17B2024.1290.3
CogVLM-17BModel Size=17B2024.0790.25
InternVL2-Llama3-76BBackbone=Llama3, Parameters=76B2024.1290
ONE-PEACE2024.1289.8
Ferret-v2-13BParameters=13B2024.1289.6
UNINEXT-HBackbone=Huge2024.1288.9
TextHawk22024.1288.2
Qwen2-VL-7BParameters=7B2024.1287.9
CoS-7BModel Size=7B, Fine-tuning protocol=LoRA2024.0787.68
InternVL2.5-8BParameters=8B2024.1287.6
UNINEXT-LModel type=Specialist2024.0486.95
G-DINOModel type=Specialist2024.0486.6
Grounding-DINO-LBackbone=Large2024.1286.6
GromaModel type=Generalist2024.0486.52
Qwen-VLModel type=Generalist, Visual Tokenizer=1.9B ViT-bigG2024.0485.83
Qwen-VL-7BModel Size=7B2024.0785.83
Ferret-13BModel Size=13B2024.0785.57
Griffon v22024.0785.42
MiniGPT-v2Model type=Generalist2024.0484.29
MiniGPTv2Fine-tuning protocol=LoRA2024.0784.29
Shikra-13BModel Size=13B2024.0783.95
FerretModel type=Generalist2024.0483.91
Ferret-7BModel Size=7B2024.0783.91
CoLAUpdate Ratio=17.2%2026.0483.4
ShikraModel type=Generalist2024.0482.93
Shikra-7BModel Size=7B2024.0782.93
Shikra-7BParameters=7B2024.1282.9
InternVL2-8BParameters=8B2024.1282.9
MDETRModel type=Specialist2024.0481.81
MDETR2024.0781.81
SwimVGUpdate Ratio=2.04%2026.0481.8
EEVGUpdate Ratio=100%2026.0481.7
HiVGUpdate Ratio=20.1%2026.0480.9
VG-LAWUpdate Ratio=100%2026.0479.7
TransVG++Update Ratio=100%2026.0478.8
MaPPERUpdate Ratio=6.2%2026.0478.7
QRNetUpdate Ratio=100%2026.0476.3
UNITER2024.0776.17
OFAModel type=Generalist2024.0472.65
OFA-LModel Size=Large2024.0772.65
TransVGUpdate Ratio=100%2026.0471.4
MAttNet2024.0768.87