List your company

Audio-Visual Captioning on DVD-Bench En

81.1AccuracyD-ORCA
61.96466.93271.976.868Feb 8, 2026

Evaluation results

MethodLinks
D-ORCANumber of Parameters=8B2026.0281.116.657.1
AVoCaDONumber of Parameters=7B2026.0272.917.9—
Ours-SFT ModelNumber of Parameters=8B2026.0271.229.831.8
Qwen3-Omni-InstructNumber of Parameters=30B-A3B2026.0267.891.3—
video-SALMONN 2+Number of Parameters=7B2026.0266.694—
ARC-Qwen-Video-NarratorNumber of Parameters=7B2026.0266.46523
Qwen2.5-OmniNumber of Parameters=7B2026.0262.783.6—