R2-D2 trains both a network and its pseudo-labels in one loss, derives an exponential relation between them, and cuts ImageNet Top-1 error from 46.50 to 41.55 with only 10% labels.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Repetitive Reprediction Deep Decipher for Semi-Supervised Learning
R2-D2 trains both a network and its pseudo-labels in one loss, derives an exponential relation between them, and cuts ImageNet Top-1 error from 46.50 to 41.55 with only 10% labels.