The 1st Winner for 5th PVUW MeViS-Text Challenge: Strong MLLMs Meet SAM3 for Referring Video Object Segmentation

TL;DR

This paper introduces a fully training-free, multimodal large language model-based pipeline that achieves state-of-the-art results in referring video object segmentation without task-specific fine-tuning.

Contribution

The authors propose a novel, training-free approach combining strong LLMs and SAM3 for video segmentation, outperforming previous methods.

Findings

Achieved first place on PVUW 2026 MeViS-Text test set

Final score of 0.909064 and J&F score of 0.7897

No task-specific fine-tuning required

Abstract

This report presents our winning solution to the 5th PVUW MeViS-Text Challenge. The track studies referring video object segmentation under motion-centric language expressions, where the model must jointly understand appearance, temporal behavior, and object interactions. To address this problem, we build a fully training-free pipeline that combines strong multimodal large language models with SAM3. Our method contains three stages. First, Gemini-3.1 Pro decomposes each target event into instance-level grounding targets, selects the frame where the target is most clearly visible, and generates a discriminative description. Second, SAM3-agent produces a precise seed mask on the selected frame, and the official SAM3 tracker propagates the mask through the whole video. Third, a refinement stage uses Qwen3.5-Plus and behavior-level verification to correct ambiguous or semantically inconsistent predictions. Without task-specific fine-tuning, our method ranks first on the PVUW 2026 MeViS-Text test set, achieving a Final score of 0.909064 and a J&F score of 0.7897. The code is available at https://github.com/Moujuruo/MeViSv2_Track_Solution_2026.