Preprint
Aug 2026
A Convolutional Layer Activation Dimensionality Reduction for Out-of-Distribution and Adversarial Attack Detection Methods
This paper carefully analyzes two state-of-the-art detection methods and their dimensionality reductions for convolutional layers and develops a novel reduction method with a controllable high-compression level.
Leandro de Souza Rosa, Lorenzo Capelli, Clara Nunes Barrancos et al.
· 0 citations