Adding a frozen FMCE convergence-score head with a tuned weight can improve image-classification accuracy by up to about 1.16 percentage points, but the paper's own equations and tables conflict.
Deep learning for safe autonomous driving: Current challenges and future directions,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
FMCE-Net++: Feature Map Convergence Evaluation and Training
Adding a frozen FMCE convergence-score head with a tuned weight can improve image-classification accuracy by up to about 1.16 percentage points, but the paper's own equations and tables conflict.