Publications related to Explainable Face Verification via Feature-Guided Gradient Backpropagation

Deep Learning Theory Through the Lens of Diagonal Linear Networks

In this PhD manuscript, we explore optimisation phenomena which occur in complex neural networks through the lens of

2

-layer diagonal linear networks. This rudimentary architecture, which consists of a two layer feedforward linear network with a diagonal ...

EPFL2024

Understanding generalization and robustness in modern deep learning

Maksym Andriushchenko

In this thesis, we study two closely related directions: robustness and generalization in modern deep learning. Deep learning models based on empirical risk minimization are known to be often non-robust to small, worst-case perturbations known as adversari ...

EPFL2024

Revisiting Character-level Adversarial Attacks for Language Models

Volkan Cevher, Grigorios Chrysos, Fanghui Liu, Yongtao Wu, Elias Abad Rocamora

Adversarial attacks in Natural Language Processing apply perturbations in the character or token levels. Token-level attacks, gaining prominence for their use of gradient-based methods, are susceptible to altering sentence semantics, leading to invalid adv ...

2024

Towards Visual Saliency Explanations of Face Verification

Touradj Ebrahimi, Yuhang Lu, Zewei Xu

In the past years, deep convolutional neural networks have been pushing the frontier of face recognition (FR) techniques in both verification and identification scenarios. Despite the high accuracy, they are often criticized for lacking explainability. The ...

2023

Discriminative Deep Feature Visualization for Explainable Face Recognition

Touradj Ebrahimi, Yuhang Lu, Zewei Xu

Despite the huge success of deep convolutional neural networks in face recognition (FR) tasks, current methods lack explainability for their predictions because of their ``black-box'' nature. In recent years, studies have been carried out to give an interp ...

2023

Hamiltonian Deep Neural Networks Guaranteeing Non-Vanishing Gradients by Design

Giancarlo Ferrari Trecate, Luca Furieri, Clara Lucía Galimberti, Liang Xu

Deep Neural Networks (DNNs) training can be difficult due to vanishing and exploding gradients during weight optimization through backpropagation. To address this problem, we propose a general class of Hamiltonian DNNs (H-DNNs) that stem from the discretiz ...

2023

Generalization Properties of NAS under Activation and Skip Connection Search

Volkan Cevher, Grigorios Chrysos, Fanghui Liu, Zhenyu Zhu

Neural Architecture Search (NAS) has fostered the automatic discovery of stateof- the-art neural architectures. Despite the progress achieved with NAS, so far there is little attention to theoretical guarantees on NAS. In this work, we study the generaliza ...

2022

Improving the Training of Compact Neural Networks for Visual Recognition

Shuxuan Guo

During the Artificial Intelligence (AI) revolution of the past decades, deep neural networks have been widely used and have achieved tremendous success in visual recognition. Unfortunately, deploying deep models is challenging because of their huge model s ...

EPFL2022

Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks

Florent Gérard Krzakala, Lenka Zdeborová, Ludovic Théo Stephan, Bruno Loureiro

Despite the non-convex optimization landscape, over-parametrized shallow networks are able to achieve global convergence under gradient descent. The picture can be radically different for narrow net-works, which tend to get stuck in badly-generalizing loca ...

2022

Approximating Relu Networks By Single-Spike Computation

Wulfram Gerstner, Stanislaw Andrzej Wozniak, Ana Stanojevic, Evangelos Eleftheriou, Giovanni Cherubini, Angeliki Pantazi

Developing energy-saving neural network models is a topic of rapidly increasing interest in the artificial intelligence community. Spiking neural networks (SNNs) are biologically inspired models that strive to leverage the energy efficiency stemming from a ...

IEEE2022

Explainable Face Verification via Feature-Guided Gradient Backpropagation

Graph Chatbot

Chat with Graph Search

Deep Learning Theory Through the Lens of Diagonal Linear Networks

Understanding generalization and robustness in modern deep learning

Revisiting Character-level Adversarial Attacks for Language Models

Towards Visual Saliency Explanations of Face Verification

Discriminative Deep Feature Visualization for Explainable Face Recognition

Hamiltonian Deep Neural Networks Guaranteeing Non-Vanishing Gradients by Design

Generalization Properties of NAS under Activation and Skip Connection Search

Improving the Training of Compact Neural Networks for Visual Recognition

Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks

Approximating Relu Networks By Single-Spike Computation

Hamiltonian Deep Neural Networks Guaranteeing Non-Vanishing Gradients by Design

Towards Visual Saliency Explanations of Face Verification

Deep Learning Theory Through the Lens of Diagonal Linear Networks

Revisiting Character-level Adversarial Attacks for Language Models

Discriminative Deep Feature Visualization for Explainable Face Recognition

Understanding generalization and robustness in modern deep learning

Improving the Training of Compact Neural Networks for Visual Recognition

Generalization Properties of NAS under Activation and Skip Connection Search

Approximating Relu Networks By Single-Spike Computation

Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks