Alexandre Araujo

发表

The Expressive Power of Deep Neural Networks with Circulant Matrices

Yann Chevaleyre, Jamal Atif, Alexandre Araujo, 2018 .

Fast & Accurate Method for Bounding the Singular Values of Convolutional Layers with Application to Lipschitz Regularization pdf

Yann Chevaleyre, Jamal Atif, Alexandre Araujo, 2020, ArXiv.

Robust Neural Networks using Randomized Adversarial Training pdf

Yann Chevaleyre, Rafael Pinot, Florian Yger, 2019, ArXiv.

On Lipschitz Regularization of Convolutional Layers using Toeplitz Matrix Theory

Yann Chevaleyre, Jamal Atif, Alexandre Araujo, 2020, AAAI.

Theoretical evidence for adversarial robustness through randomization: the case of the Exponential family pdf

Hisashi Kashima, Rafael Pinot, Florian Yger, 2019, ArXiv.

On the Expressive Power of Deep Fully Circulant Neural Networks pdf

Yann Chevaleyre, Jamal Atif, Benjamin Négrevergne, 2019, ArXiv.

Training compact deep learning models for video classification using circulant matrices pdf

Yann Chevaleyre, Jamal Atif, Benjamin Négrevergne, 2018, ECCV Workshops.

Understanding and Training Deep Diagonal Circulant Neural Networks

Yann Chevaleyre, Jamal Atif, Alexandre Araujo, 2020, ECAI.

Building Compact and Robust Deep Neural Networks with Toeplitz Matrices pdf

Alexandre Araujo, 2021, ArXiv.

Scalable Lipschitz Residual Networks with Convex Potential Flows pdf

Alexandre Allauzen, Alexandre Araujo, Laurent Meunier, 2021, ArXiv.