# Decoupled Seg Tokens Make Stronger Reasoning Video Segmenter and Grounder

Source: [Decoupled Seg Tokens Make Stronger Reasoning Video Segmenter and Grounder](<https://www.sota2.com/research/paper/68fa79af1e209131cea7edfb>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:57:18.504Z

Authors: Dang Jisheng, Wu Xudong, Wang Bimei, Lv Ning, Chen Jiayu, Jingwen Zhao, Yichu liu, Jizhao Liu, Juncheng Li, Teng Wang (7) \_\_INSTITUTION\_10\_\_ Lanzhou University, (2) National University of Singapore, (3) Sun Yat-sen University, (4) Jinan University, (5) South China University of Technology, (6) Zhejiang University, (7) The University of Hong Kong )

Published: 2025-06-28T13:30:36.000Z

## Abstract

Existing video segmenter and grounder approaches, exemplified by Sa2VA, directly fuse features within segmentation models. This often results in an undesirable entanglement of dynamic visual information and static semantics, thereby degrading segmentation accuracy. To systematically mitigate this issue, we propose DeSa2VA, a decoupling-enhanced prompting scheme integrating text pre-training and a linear decoupling module to address the information processing limitations inherent in SAM-2. Specifically, first, we devise a pre-training paradigm that converts textual ground-truth labels into point-level prompts while generating corresponding text masks. These masks are refined through a hybrid loss function to strengthen the model's semantic grounding capabilities. Next, we employ linear projection to disentangle hidden states that generated by a large language model into distinct textual and visual feature subspaces. Finally, a dynamic mask fusion strategy synergistically combines these decoupled features through triple supervision from predicted text/visual masks and ground-truth annotations. Extensive experiments demonstrate state-of-the-art performance across diverse tasks, including image segmentation, image question answering, video segmentation, and video question answering. Our codes are available at https://github.com/longmalongma/DeSa2VA.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2506.22880>)

[PDF](<https://arxiv.org/pdf/2506.22880.pdf>)