Patrick O'Reilly

I'm a doctoral student in the Department of Computer Science at Northwestern University and a member of the Interactive Audio Lab. I received a BA in Mathematics and Music from Carleton College and an MS in Computer Science from the University of Illinois at Chicago. My research interests include adversarial robustness for audio interfaces, music information retrieval, and machine learning techniques for controllable audio generation.


Patrick O'Reilly, Pranjal Awasthi, Aravindan Vijayaraghavan, Bryan Pardo

Paper (Submitted to ICASSP 2022) · Project Page · Talk @ BishBash

We demonstrate a novel audio-domain adversarial attack that modifies benign audio using an interpretable and differentiable parametric transformation - adaptive filtering. Unlike existing state-of-the-art attacks, our proposed method does not require a complex optimization procedure or generative model, relying only on a simple variant of gradient descent to tune filter parameters. We demonstrate the effectiveness of our method by performing over-the-air attacks against a state-of-the-art speaker verification model. Our results demonstrate the potential of transformations beyond direct waveform addition for concealing high-magnitude adversarial perturbations, allowing adversaries to attack more effectively in challenging real-world settings.

Ethan Manilow, Patrick O'Reilly, Prem Seetharaman, Bryan Pardo

Paper (Submitted to ICASSP 2022) · Project Page · Code

We showcase an unsupervised method that repurposes deep models trained for music generation and music tagging for audio source separation, without any retraining. An audio generation model is conditioned on an input mixture, producing a latent encoding of the audio used to generate audio. This generated audio is fed to a pretrained music tagger that creates source labels. The cross-entropy loss between the tag distribution for the generated audio and a predefined distribution for an isolated source is used to guide gradient ascent in the (unchanging) latent space of the generative model. This work points to the vast and heretofore untapped potential of large pretrained music models for audio-to-audio tasks like source separation.

From June 2019 through September 2020, I worked on prototyping lightweight systems for vehicle state recognition and intrusion detection as part of a collaborative research project between Caterpillar Inc. and the UIC Innovation Center. Both projects focused on machine-learning methods for sensor data classification, with the goal of providing actionable insights to machinery operators in real time. Additional responsibilities included data acquisition and working with Caterpillar to build and deploy a software solution on proprietary hardware.