List your company

SlideASR on SlideSpeech (test)

10.31WERVAPO-7B
2.307656.3238110.34164.3562Oct 8, 2025

Evaluation results

MethodLinks
VAPO-7BContext usage protocol=Slide image as context (End-to-End)2025.1010.3110.842.8797.32—
VAPO-3BContext usage protocol=Slide image as context (End-to-End)2025.1010.7311.243.5596.57—
MaLa-ASRContext usage protocol=Slide text as context (Pipeline)2025.1011.2611.527.6792.5—
Qwen3-Omni-30B-A3BContext usage protocol=Contextless2025.1011.7112.214.6495.5—
Qwen2.5-Omni-7BContext usage protocol=Contextless2025.1011.7512.25.3994.78—
Qwen2.5-Omni-3BContext usage protocol=Contextless2025.101212.455.7294.41—
Qwen2-AudioContext usage protocol=Contextless2025.1013.1913.597.5392.91—
Qwen2.5-Omni-7BContext usage protocol=Slide image as context (End-to-End)2025.1014.9715.586.3393.99—
MiniCPM-o-2.6Context usage protocol=Contextless2025.1018.7119.418.991.5—
LCB-netContext usage protocol=Slide text as context (Pipeline)2025.1019.2118.8923.776.48—
Qwen2.5-Omni-3BContext usage protocol=Slide image as context (End-to-End)2025.1019.9920.719.894.44—
Qwen3-Omni-30B-A3BContext usage protocol=Slide image as context (End-to-End)2025.1024.1324.8813.4494.74—
Mi-DashengContext usage protocol=Slide text as context (Pipeline)2025.1047.2149.3417.2191—
Qwen3-Omni-30B-A3BContext usage protocol=Slide text as context (Pipeline)2025.1057.1259.2726.7596.34—
Qwen2-AudioContext usage protocol=Slide text as context (Pipeline)2025.10146.08152.4156.9988.98—
MiniCPM-o-2.6Context usage protocol=Slide image as context (End-to-End)2025.10210.37220.9660.9283.22—
Megrez-Omni2025.10————44.9
MiniCPM-o-2.62025.10————63.28
Qwen2.5-Omni-3BParameters=3B2025.10————16.54
Qwen2.5-Omni-7BParameters=7B2025.10————12.87