| title | MFVSR: A Multi-Style and Fine-Grained Benchmark for Video Subtitle Removal |
|---|---|
| authors | Xiyan Huang, Qingdong He, Peng Tang, Bo Yin, Chaoyi Wang, Xiaobin Hu |
| date | 2026-10-13 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2610.18461 |
| upvotes | 18 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Video subtitle removal (VSR) is an important video restoration task, yet existing evaluation remains limited by low-resolution datasets, fragmented test sets, and generic video-inpainting protocols that fail to adequately capture localized restoration errors, unintended modifications, temporal instability, and incomplete outputs. In this paper, we introduce MFVSR, a large-scale multi-style benchmark for fine-grained and reliability-aware video subtitle removal evaluation. MFVSR distinguishes itself through three key features: (1) Large-scale and diverse data, comprising 3,000 paired videos from Real-HQ, Real-LQ, and AI-Generated sources, with an auxiliary set of 953 in-the-wild real videos for real-world coverage analysis; (2) Controlled subtitle diversity, spanning 16 languages and four factors covering spatial layout, temporal behavior, visual style, and language mode; (3) Fine-grained evaluation, introducing an information-conditioned, spatially decoupled protocol that separates blind from mask-guided removal and measures Restoration Fidelity, Preservation Fidelity, Temporal Fidelity, coverage, and failures. Experiments on six representative methods reveal substantial discrepancies between local restoration, outside-region preservation, temporal consistency, and full-set reliability that are obscured by conventional full-frame metrics. MFVSR therefore provides a more comprehensive and diagnostic foundation for evaluating and advancing video subtitle removal methods. |
cite
@misc{huang2026mfvsr,
title={MFVSR: A Multi-Style and Fine-Grained Benchmark for Video Subtitle Removal},
author={Xiyan Huang and Qingdong He and Peng Tang and Bo Yin and Chaoyi Wang and Xiaobin Hu},
year={2026},
eprint={2610.18461},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2610.18461}
}