CDMAD addresses biased classifiers in SSL by refining pseudo-labels and class predictions, improving representation quality. Extensive experiments show CDMAD's effectiveness across various scenarios. It ensures Fisher consistency for balanced error.
Til et andet sprog
fra kildeindhold
arxiv.org
Vigtigste indsigter udtrukket fra
by Hyuck Lee,He... kl. arxiv.org 03-18-2024
https://arxiv.org/pdf/2403.10391.pdfDybere Forespørgsler