A weighted sum of forward and reverse KL divergence with two temperatures is proposed as a logit-based knowledge distillation loss, with reported gains on image classification and semantic segmentation.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Balance Divergence for Knowledge Distillation
A weighted sum of forward and reverse KL divergence with two temperatures is proposed as a logit-based knowledge distillation loss, with reported gains on image classification and semantic segmentation.