EUVIMEDEUVIMED entdecken
Crossref

Learning Beyond Vision: Vision-Language Distillation and Edge-Aware Mix Diffusion in Semi-Supervised Semantic Segmentation

Rui Yang, Yunfei Bai, Yuehua Liu, Xiaomao Li, Shaorong Xie

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (14) · 11676-11684

DOI: 10.1609/aaai.v40i14.38152 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Rui Yang und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In semi-supervised semantic segmentation (SSSS), segmentation performance is heavily constrained by the quality of pseudo labels. However, prevalent pseudo-label optimization approaches rely on the model’s internal self-correction. When the model fails to recognize or adequately represent certain classes, this self-enhancement mechanism amplifies initial mistakes, ultimately leading to poor semantic or spatial consistency. To address this limitation, we propose ViLaDiff to enhance pseudo-label quality. Specifically, ViLaDiff first employs a prompt-guided image captioning task to generate descriptive text for each input image, providing high-level semantic context. To our knowledge, this is the first attempt to introduce vision-language modeling into SSSS. We design a vision-language fusion module to enhance feature semantics and discriminative capability. It integrates cross-modal interactions with dual-path knowledge to ensure semantic consistency. Additionally, while language provides high-level semantic guidance, it is inherently limited in expressing fine-grained spatial structures. Therefore, we propose an edge-aware mixed-noise diffusion process. It simulates feature-level uncertainty through Gaussian perturbations and introduces class-flipping noise into the masks to model misclassification errors. To enhance boundary refinement, we apply a higher flipping probability along mask edges, enabling edge-aware modeling during denoising. Extensive experiments on public benchmarks validate that our method significantly improves pseudo-label quality and segmentation performance.

Zitieren

APA

Yang, R., Bai, Y., Liu, Y., Li, X., & Xie, S. (2026). Learning Beyond Vision: Vision-Language Distillation and Edge-Aware Mix Diffusion in Semi-Supervised Semantic Segmentation. Proceedings of the AAAI Conference on Artificial Intelligence, 40(14), 11676-11684. https://doi.org/10.1609/aaai.v40i14.38152

Vancouver

Yang R, Bai Y, Liu Y, Li X, Xie S. Learning Beyond Vision: Vision-Language Distillation and Edge-Aware Mix Diffusion in Semi-Supervised Semantic Segmentation. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(14):11676-11684. doi:10.1609/aaai.v40i14.38152

BibTeX
@article{Yang2026,
  author = {Yang, Rui and Bai, Yunfei and Liu, Yuehua and Li, Xiaomao and Xie, Shaorong},
  title = {Learning Beyond Vision: Vision-Language Distillation and Edge-Aware Mix Diffusion in Semi-Supervised Semantic Segmentation},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {14},
  pages = {11676--11684},
  doi = {10.1609/aaai.v40i14.38152}
}
Kostenlos starten →