Publications (21)
The LogBarrier adversarial attack: making effective use of decision boundary information
Chris Finlay, Aram-Alexandre Pooladian, Adam M. Oberman
Adversarial attacks for image classification are small perturbations to images that are designed to cause misclassification by a model. Adversarial attacks formally correspond to a…
Improved accuracy of monotone finite difference schemes on point clouds and regular grids
Chris Finlay, Adam Oberman
Finite difference schemes are the method of choice for solving nonlinear, degenerate elliptic PDEs, because the Barles-Sougandis convergence framework [Barles and Sougandidis, Asym…
Hydromagnetic quasi-geostrophic modes in rapidly rotating planetary cores
Elisabeth Canet, Chris Finlay, Alexandre Fournier
The core of a terrestrial-type planet consists of a spherical shell of rapidly rotating, electrically conducting, fluid. Such a body supports two distinct classes of quasi-geostrop…
Climate & BCG: Effects on COVID-19 Death Growth Rates
Chris Finlay, Bruce A. Bassett
Multiple studies have suggested the spread of COVID-19 is affected by factors such as climate, BCG vaccinations, pollution and blood type. We perform a joint study of these factors…
Adversarial Boot Camp: label free certified robustness in one epoch
Ryan Campbell, Chris Finlay, Adam M Oberman
Machine learning models are vulnerable to adversarial attacks. One approach to addressing this vulnerability is certification, which focuses on models that are guaranteed to be rob…
Approximate homogenization of convex nonlinear elliptic PDEs
Chris Finlay, Adam M. Oberman
We approximate the homogenization of fully nonlinear, convex, uniformly elliptic Partial Differential Equations in the periodic setting, using a variational formula for the optimal…
Scaleable input gradient regularization for adversarial robustness
Chris Finlay, Adam M Oberman
In this work we revisit gradient regularization for adversarial robustness with some new ingredients. First, we derive new per-image theoretical robustness bounds based on local gr…
Improved robustness to adversarial examples using Lipschitz regularization of the loss
Chris Finlay, Adam Oberman, Bilal Abbasi
We augment adversarial training (AT) with worst case adversarial training (WCAT) which improves adversarial robustness by 11% over the current state-of-the-art result in the $\ell_…
Multi-Resolution Continuous Normalizing Flows
Vikram Voleti, Chris Finlay, Adam Oberman +1
Recent work has shown that Neural Ordinary Differential Equations (ODEs) can serve as generative models of images using the perspective of Continuous Normalizing Flows (CNFs). Such…
Approximate homogenization of fully nonlinear elliptic PDEs: estimates and numerical results for Pucci type equations
Chris Finlay, Adam M. Oberman
We are interested in the shape of the homogenized operator for PDEs which have the structure of a nonlinear Pucci operator. A typical operator is $H^{a_1,a_2}(Q,x)…
TABASCAL II: Removing Multi-Satellite Interference from Point-Source Radio Astronomy Observations
Chris Finlay, Bruce A. Bassett, Martin Kunz +1
In the first TABASCAL paper we showed how to calibrate in the presence of Radio Frequency Interference (RFI) sources by simultaneously isolating the trajectories and signals of the…
Deep learning approach for identification of HII regions during reionization in 21-cm observations -- III. image recovery
Michele Bianco, Sambit. K. Giri, Rohit Sharma +7
The low-frequency component of the upcoming Square Kilometre Array Observatory (SKA-Low) will be sensitive enough to construct 3D tomographic images of the 21-cm signal distributio…
A principled approach for generating adversarial images under non-smooth dissimilarity metrics
Aram-Alexandre Pooladian, Chris Finlay, Tim Hoheisel +1
Deep neural networks perform well on real world data but are prone to adversarial perturbations: small changes in the input easily lead to misclassification. In this work, we propo…
Learning normalizing flows from Entropy-Kantorovich potentials
Chris Finlay, Augusto Gerolin, Adam M Oberman +1
We approach the problem of learning continuous normalizing flows from a dual perspective motivated by entropy-regularized optimal transport, in which continuous normalizing flows a…
Trajectory Based RFI Subtraction and Calibration for Radio Interferometry
Chris Finlay, Bruce A. Bassett, Martin Kunz +1
Radio interferometry calibration and Radio Frequency Interference (RFI) removal are usually done separately. Here we show that jointly modelling the antenna gains and RFI has signi…
Deterministic Gaussian Averaged Neural Networks
Ryan Campbell, Chris Finlay, Adam M Oberman
We present a deterministic method to compute the Gaussian average of neural networks used in regression and classification. Our method is based on an equivalence between training w…
Deep Learning improves identification of Radio Frequency Interference
Alireza Vafaei Sadr, Bruce A. Bassett, Nadeem Oozeer +2
Flagging of Radio Frequency Interference (RFI) is an increasingly important challenge in radio astronomy. We present R-Net, a deep convolutional ResNet architecture that significan…
Lipschitz regularized Deep Neural Networks generalize and are adversarially robust
Chris Finlay, Jeff Calder, Bilal Abbasi +1
In this work we study input gradient regularization of deep neural networks, and demonstrate that such regularization leads to generalization proofs and improved adversarial robust…
Farkas layers: don't shift the data, fix the geometry
Aram-Alexandre Pooladian, Chris Finlay, Adam M Oberman
Successfully training deep neural networks often requires either batch normalization, appropriate weight initialization, both of which come with their own challenges. We propose an…
Calibrated Top-1 Uncertainty estimates for classification by score based models
Adam M. Oberman, Chris Finlay, Alexander Iannantuono +1
While the accuracy of modern deep learning models has significantly improved in recent years, the ability of these models to generate uncertainty estimates has not progressed to th…
How to train your neural ODE: the world of Jacobian and kinetic regularization
Chris Finlay, Jörn-Henrik Jacobsen, Levon Nurbekyan +1
Training neural ODEs on large datasets has not been tractable due to the necessity of allowing the adaptive numerical ODE solver to refine its step size to very small values. In pr…