ReVEAL: Eliciting Role-Aware Latent Visual Evidence from MLLMs for Reasoning Video Object Segmentation

Benchmark Model Rank Results
referring-video-object-segmentation-on-ref-davis17ReVEAL (Qwen3VL-8B)#2J&F: 80.7
referring-video-object-segmentation-on-ref-davis17ReVEAL (Qwen2.5VL-7B)#8J&F: 77.7