List your company

Fine-grained Visual Question Answering on V*Bench

92.15Overall AccuracyFOCUS
48.584459.894771.20582.5153Jun 26, 2025

Evaluation results

MethodLinks
FOCUSBackbone=LLaVA-OV-7B, Steps=4, Overrun=true2025.0692.1593.9189.477.63
FOCUSBackbone=LLaVA-OV-7B, Steps=6, Overrun=true2025.0692.1593.9189.4710.22
FOCUSBackbone=LLaVA-OV-7B, Steps=2, Overrun=true2025.0691.6293.9388.155.16
FOCUSBackbone=LLaVA-OV-7B, Steps=3, Overrun=true2025.0691.6293.9188.156.37
FOCUSBackbone=LLaVA-OV-7B, Steps=8, Overrun=true2025.0691.6293.0489.4712.84
ZoomEyeBackbone=LLaVA-OV-7B, Depth=5, Number of crops=42025.0691.193.9186.8423.98
FOCUSBackbone=LLaVA-OV-7B, Steps=1, Overrun=true2025.0690.5793.0486.844.05
ZoomEyeBackbone=LLaVA-OV-7B, Depth=2, Number of crops=42025.0690.0593.0485.5321.08
FOCUSBackbone=LLaVA-OV-7B, Steps=2, Overrun=false2025.0689.0192.1784.214.23
FOCUSBackbone=LLaVA-OV-7B, Steps=1, Overrun=false2025.0683.2487.8276.311.47
ZoomEyeBackbone=LLaVA-OV-7B, Depth=1, Number of crops=42025.0681.1581.7480.2613.34
ZoomEyeBackbone=LLaVA-1.5-7B, Depth=5, Number of crops=42025.0677.4880.8772.3748.63
ZoomEyeBackbone=LLaVA-OV-7B, Depth=1, Number of crops=22025.0675.9277.3973.6810.52
SEAL2025.0673.68——25.53
FOCUSBackbone=LLaVA-1.5-7B, Steps=8, Overrun=true2025.0672.7772.1773.6813.28
ZoomEyeBackbone=LLaVA-1.5-7B, Depth=2, Number of crops=42025.0671.267.8376.3244.54
FOCUSBackbone=LLaVA-1.5-7B, Steps=6, Overrun=true2025.0670.6870.4371.0510.71
FOCUSBackbone=LLaVA-1.5-7B, Steps=4, Overrun=true2025.0668.0666.9669.748.27
FOCUSBackbone=LLaVA-1.5-7B, Steps=3, Overrun=true2025.0667.0166.0968.426.79
FOCUSBackbone=LLaVA-1.5-7B, Steps=2, Overrun=true2025.0666.4966.0967.115.7
FOCUSBackbone=LLaVA-1.5-7B, Steps=1, Overrun=true2025.0664.463.4865.794.86
ViCropBackbone=LLaVA-1.5-7B, Mode=rel-att-high2025.0659.1658.2660.5312.26
FOCUSBackbone=LLaVA-1.5-7B, Steps=2, Overrun=false2025.0657.0753.9161.844.25
DC2Backbone=LLaVA-v.1.5-7B2025.0657——18.18
ViCropBackbone=LLaVA-1.5-7B, Mode=grad-att-high2025.0654.9753.0457.96.63
FOCUSBackbone=LLaVA-1.5-7B, Steps=1, Overrun=false2025.0651.346.9557.891.47
ZoomEyeBackbone=LLaVA-1.5-7B, Depth=1, Number of crops=42025.0650.7841.7464.4720.37
ZoomEyeBackbone=LLaVA-1.5-7B, Depth=1, Number of crops=22025.0650.2641.7463.1612.5