EUVIMEDEUVIMED entdecken
Crossref

Soft-Labelling for Budget-Constrained Semantic Segmentation: Bringing Coherence to label Down-Sampling

Roberto Alcover-Couso, Marcos Escudero-Viñolo, Juan C. SanMiguel, José M. Martinez

Computational Visual Media · 2026-06 · 12 (3) · 765-785

DOI: 10.26599/cvm.2025.9450470 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computational Visual Media, 2026-06. Verfasst von Roberto Alcover-Couso und 3 weiteren Autor:innen. Die Arbeit nennt 79 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In semantic segmentation, training data down-sampling is commonly performed due to limited resources, the need to adapt image size to the model input, or improve data augmentation. This down-sampling typically employs different strategies for the image data and the annotated labels. Such discrepancy leads to mismatches between the down-sampled color and label images. Hence, the training performance significantly decreases as the down-sampling factor increases. In this paper, we bring together the down-sampling strategies for the image data and the training labels. To that aim, we propose a novel framework for label down-sampling via soft-labeling that better conserves label information after down-sampling. Therefore, fully aligning soft-labels with image data to keep the distribution of the sampled pixels. This proposal also produces reliable annotations for under-represented semantic classes. Altogether, it allows training competitive models at lower resolutions. Experiments show that the proposal outperforms other down-sampling strategies. Moreover, state-of-the-art performance is achieved for reference benchmarks, but employing significantly less computational resources than foremost approaches. This proposal enables competitive research for semantic segmentation under resource constraints.

Zitieren

APA

Alcover-Couso, R., Escudero-Viñolo, M., SanMiguel, J. C., & Martinez, J. M. (2026). Soft-Labelling for Budget-Constrained Semantic Segmentation: Bringing Coherence to label Down-Sampling. Computational Visual Media, 12(3), 765-785. https://doi.org/10.26599/cvm.2025.9450470

Vancouver

Alcover-Couso R, Escudero-Viñolo M, SanMiguel JC, Martinez JM. Soft-Labelling for Budget-Constrained Semantic Segmentation: Bringing Coherence to label Down-Sampling. Computational Visual Media. 2026;12(3):765-785. doi:10.26599/cvm.2025.9450470

BibTeX
@article{AlcoverCouso2026,
  author = {Alcover-Couso, Roberto and Escudero-Viñolo, Marcos and SanMiguel, Juan C. and Martinez, José M.},
  title = {Soft-Labelling for Budget-Constrained Semantic Segmentation: Bringing Coherence to label Down-Sampling},
  journal = {Computational Visual Media},
  year = {2026},
  volume = {12},
  number = {3},
  pages = {765--785},
  doi = {10.26599/cvm.2025.9450470}
}
Kostenlos starten →