Metraipolis › archive › dc_04ef9876540a
T2LSC-Bench: Benchmarking Localized Semantic Control in Text-to-Image Generation — arXiv:2609.02255
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02255
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T05:49:57.665Z
- SHA-256 of the archived text
bb7887a422fe4432352475b1aef9aeba3e4f234a86a19d8fa25f572a9c3354bd- Stable URL of this page
https://metraipolis.com/documento/dc_04ef9876540a
How to cite
T2LSC-Bench: Benchmarking Localized Semantic Control in Text-to-Image Generation — arXiv:2609.02255 — arXiv (en), https://arxiv.org/abs/2609.02255 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_04ef9876540a (Metraipolis archive dc_04ef9876540a, sha256 bb7887a422fe4432352475b1aef9aeba3e4f234a86a19d8fa25f572a9c3354bd)
BibTeX
@misc{dc04ef9876540a,
title = {T2LSC-Bench: Benchmarking Localized Semantic Control in Text-to-Image Generation — arXiv:2609.02255},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02255},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_04ef9876540a; sha256 bb7887a422fe4432352475b1aef9aeba3e4f234a86a19d8fa25f572a9c3354bd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02255. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recent text-to-image models have become increasingly capable of rendering explicit text, but reliable localized text control requires more than generating the correct string. In applications such as product labeling, signage, and interface design, target text should be rendered within a designated text-bearing region without altering the predefined subject identity or surrounding scene semantics. We refer to violations of this requirement as target-text-associated semantic leakage, in which target-text semantics are expressed through non-textual visual content beyond the designated anchor. Existing visual-text benchmarks primarily evaluate readability, spelling accuracy, and layout, leaving this form of semantic leakage largely unexamined. We introduce T2LSC-Bench, a controlled diagnostic benchmark comprising 50 seed subjects and 1,200 prompt cases per model, yielding 7,160 evaluated images across six models. Its factorized design varies semantic relation, scene openness, prompt mode, and language. A dual-branch protocol combines OCR-VLM text verification with structured VLM semantic judgments to measure Text-at-Anchor Accuracy (TAA), Semantic Subject Preservation (SSP), Semantic Leakage Rate (SLR), and Conditional Semantic Leakage Rate (cSLR). Under stress-test conditions, SLR increases from 1.2% to 18.1% and cSLR from 1.3% to 18.2%, whereas TAA decreases only from 91.4% to 90.9%. Anti-leakage prompting reduces SLR from 16.6% to 8.4% without degrading rendering accuracy. Human validation on 420 images shows strong agreement between automatic and adjudicated annotations. These results show that accurate text rendering does not guarantee local containment of target-text semantics. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-03 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2609.02255 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T05:25:58.087Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).