DEIG: Detail-Enhanced Instance Generation with Fine-Grained Semantic Control
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (5) · 3677-3685
DOI: 10.1609/aaai.v40i5.37367 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Shiyan Du und 3 weiteren Autor:innen.
Abstract (Originalsprache)
Multi-Instance Generation has advanced significantly in spatial placement and attribute binding. However, existing approaches still face challenges in fine-grained semantic understanding, particularly when dealing with complex textual descriptions.To overcome these limitations, we propose DEIG, a novel framework for fine-grained and controllable multi-instance generation. DEIG integrates an instance Detail Extractor (IDE) that transforms text encoder embeddings into compact, instance-aware representations, and a Detail Fusion Module (DFM) that applies instance-based masked attention to prevent attribute leakage across instances. These components enable DEIG to generate visually coherent multi-instance scenes that precisely match rich, localized textual descriptions. To support fine-grained supervision, we construct a high-quality dataset with detailed, compositional instance captions generated by VLMs. We also introduce DEIG-Bench, a new benchmark with region-level annotations and multi-attribute prompts for both humans and objects.Experiments demonstrate that DEIG consistently outperforms existing approaches across multiple benchmarks in spatial consistency, semantic accuracy, and compositional generalization. Moreover, DEIG functions as a plug-and-play module, making it easily integrable into standard diffusion-based pipelines.
Zitieren
APA
Du, S., Yue, C., Cheng, X., & Zhang, D. (2026). DEIG: Detail-Enhanced Instance Generation with Fine-Grained Semantic Control. Proceedings of the AAAI Conference on Artificial Intelligence, 40(5), 3677-3685. https://doi.org/10.1609/aaai.v40i5.37367
Vancouver
Du S, Yue C, Cheng X, Zhang D. DEIG: Detail-Enhanced Instance Generation with Fine-Grained Semantic Control. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(5):3677-3685. doi:10.1609/aaai.v40i5.37367
BibTeX
@article{Du2026,
author = {Du, Shiyan and Yue, Conghan and Cheng, Xinyu and Zhang, Dongyu},
title = {DEIG: Detail-Enhanced Instance Generation with Fine-Grained Semantic Control},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {5},
pages = {3677--3685},
doi = {10.1609/aaai.v40i5.37367}
}
EUVIMED