# Audit After Segmentation: Reference-Free Mask Quality Assessment for Language-Referred Audio-Visual Segmentation

Source: [Audit After Segmentation: Reference-Free Mask Quality Assessment for Language-Referred Audio-Visual Segmentation](<https://www.sota2.com/research/paper/698423f3a8b01e76cc8ee63a>)

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

Data updated: 2026-02-06T05:00:04.454Z

Authors: Jinxing Zhou, Yanghao Zhou, Yaoting Wang, Zongyan Han, Jiaqi Ma, Henghui Ding, Rao Muhammad Anwer, Hisham Cholakkal

Published: 2026-02-03T07:47:59.000Z

## Abstract

Language-referred audio-visual segmentation (Ref-AVS) aims to segment target objects described by natural language by jointly reasoning over video, audio, and text. Beyond generating segmentation masks, providing rich and interpretable diagnoses of mask quality remains largely underexplored. In this work, we introduce Mask Quality Assessment in the Ref-AVS context (MQA-RefAVS), a new task that evaluates the quality of candidate segmentation masks without relying on ground-truth annotations as references at inference time. Given audio-visual-language inputs and each provided segmentation mask, the task requires estimating its IoU with the unobserved ground truth, identifying the corresponding error type, and recommending an actionable quality-control decision. To support this task, we construct MQ-RAVSBench, a benchmark featuring diverse and representative mask error modes that span both geometric and semantic issues. We further propose MQ-Auditor, a multimodal large language model (MLLM)-based auditor that explicitly reasons over multimodal cues and mask information to produce quantitative and qualitative mask quality assessments. Extensive experiments demonstrate that MQ-Auditor outperforms strong open-source and commercial MLLMs and can be integrated with existing Ref-AVS systems to detect segmentation failures and support downstream segmentation improvement. Data and codes will be released at https://github.com/jasongief/MQA-RefAVS.

## 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.

- [Image-based Audio-Visual Segmentation Quality Assessment on MQ-RAVSBench (Unseen)](<https://www.sota2.com/research/sota/image-based-audio-visual-segmentation-quality-assessment-on-mq-ravsbench-unseen.md>): Image-based Audio-Visual Segmentation Quality Assessment; MQ-RAVSBench (Unseen); RMSE (Perfect): 0.274.

- [Image-based Audio-Visual Segmentation Quality Assessment on MQ-RAVSBench (Seen)](<https://www.sota2.com/research/sota/image-based-audio-visual-segmentation-quality-assessment-on-mq-ravsbench-seen.md>): Image-based Audio-Visual Segmentation Quality Assessment; MQ-RAVSBench (Seen); RMSE (Perfect): 0.359.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2602.03892>)

[PDF](<https://arxiv.org/pdf/2602.03892.pdf>)
