EUVIMEDEUVIMED entdecken
Crossref

Appearance-Motion Decomposed Alignment for Text-Video Retrieval

Meng Meng, Zichang Tan, Yong Zhang, Xu Zhou

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (10) · 7999-8007

DOI: 10.1609/aaai.v40i10.37745 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Meng Meng und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Text-video retrieval aims to bridge vision and language areas, which is a crucial task in multi-modal intelligence. The core idea is to learn video and textual features to quantify their semantic relevance. A common limitation in current approaches is the oversimplification of video content, where complex spatiotemporal structures are compressed into a single global representation. Consequently, these methods struggle to fully capture dynamic visual variations and discriminative appearance inside a video, further complicating cross-modal alignment. To alleviate these issues, we introduce a novel decoupling approach that independently processes appearance and motion cues, capitalizing on their complementary nature for more expressive video modeling. Specifically, we propose an appearance-motion decomposed network (AMD-Net) to decouple spatial-level appearance and temporal-level motion understanding via the discriminative appearance learning and multi-scale motion learning modules. The proposed model enjoys several merits. First, the designed discriminative appearance learning module with a Singular Value Decomposition (SVD) based prototype initialization can effectively reduce redundant appearance information, and a high-order cross-aggregation mechanism enhances prototype resilience and facilitates comprehensive video understanding. Second, the proposed multi-scale motion learning (MML) module can capture motion features at varying temporal scales, which are complementary to appearance features for accurate text-video retrieval. Extensive experiments on five standard benchmarks demonstrate that our method performs favorably against state-of-the-art methods.

Zitieren

APA

Meng, M., Tan, Z., Zhang, Y., & Zhou, X. (2026). Appearance-Motion Decomposed Alignment for Text-Video Retrieval. Proceedings of the AAAI Conference on Artificial Intelligence, 40(10), 7999-8007. https://doi.org/10.1609/aaai.v40i10.37745

Vancouver

Meng M, Tan Z, Zhang Y, Zhou X. Appearance-Motion Decomposed Alignment for Text-Video Retrieval. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(10):7999-8007. doi:10.1609/aaai.v40i10.37745

BibTeX
@article{Meng2026,
  author = {Meng, Meng and Tan, Zichang and Zhang, Yong and Zhou, Xu},
  title = {Appearance-Motion Decomposed Alignment for Text-Video Retrieval},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {10},
  pages = {7999--8007},
  doi = {10.1609/aaai.v40i10.37745}
}
Kostenlos starten →