A kernel-matrix SVD followed by average pooling gives a controllable, high-compression representation of convolutional activations that preserves out-of-distribution and adversarial-attack detection performance.
cc/paper_files/paper/2021/file/3941c4358616274ac2436eacf67fae05-Paper.pdf
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A Convolutional Layer Activation Dimensionality Reduction for Out-of-Distribution and Adversarial Attack Detection Methods
A kernel-matrix SVD followed by average pooling gives a controllable, high-compression representation of convolutional activations that preserves out-of-distribution and adversarial-attack detection performance.