Search Results - "Luzi, Lorenzo"

Overparameterization and Double Descent in PCA, GANs, and Diffusion Models by Luzi, Lorenzo

“…This PhD thesis constitutes a synthesis of my doctoral work, which addresses various aspects of study related to generative modeling with a particular focus on…”
Get full text

Awake prone positioning for patients with COVID-19-related respiratory failure: a systematic review and meta-analysis by Graziani, Mara, Rigutini, Andrea Galeazzo, Bartolini, Diletta, Traballi, Laura, Luzi, Lorenzo, Regina, Rossana, Bossi, Francesco, Caponi, Carla, Becattini, Cecilia

“…The role of awake prone positioning (aPP) in patients with acute hypoxemic respiratory failure is debated. We performed a systematic review and meta-analysis…”
Get full text

Memorization in Generative Networks and Mixtures of GANs by Luzi, Lorenzo

“…We demonstrate that memorization (perfectly fitting the training data) is necessary to avoid mode collapse in generative networks. Using a straightforward…”
Get full text

Safety and effectiveness of oral anticoagulants in patients with atrial fibrillation and stage 4 chronic kidney disease: a real-world experience by Talerico, Rosa, Brando, Elisa, Luzi, Lorenzo, Vedovati, Maria Cristina, Giustozzi, Michela, Verso, Melina, Di Gennaro, Leonardo, Basso, Maria, Ferretti, Antonietta, Porfidia, Angelo, De Candia, Erica, Pola, Roberto, Agnelli, Giancarlo, Becattini, Cecilia

“…It is still uncertain whether direct oral anticoagulants (DOACs) perform better than vitamin K antagonists (VKAs) in subjects with non-valvular atrial…”
Get full text

Titan: Bringing the Deep Image Prior to Implicit Representations by Luzi, Lorenzo, LeJeune, Daniel, Siahkoohi, Ali, Alemohammad, Sina, Saragadam, Vishwanath, Babaei, Hossein, Liu, Naiming, Wang, Zichao, Baraniuk, Richard G.

“…We study the interpolation capabilities of implicit neural representations (INRs) of images. In principle, INRs promise a number of advantages, such as…”
Get full text

Evaluating generative networks using Gaussian mixtures of image features by Luzi, Lorenzo, Marrero, Carlos Ortiz, Wynar, Nile, Baraniuk, Richard G., Henry, Michael J.

“…We develop a measure for evaluating the performance of generative networks given two sets of images. A popular performance measure currently used to do this is…”
Get full text

Acoustic firearm discharge detection and classification in an enclosed environment by Luzi, Lorenzo, Gonzalez, Eric, Bruillard, Paul, Prowant, Matthew, Skorpik, James, Hughes, Michael, Child, Scott, Kist, Duane, McCarthy, John E.

“…Two different signal processing algorithms are described for detection and classification of acoustic signals generated by firearm discharges in small enclosed…”
Get full text

NFT-K: Non-Fungible Tangent Kernels by Alemohammad, Sina, Babaei, Hossein, Barberan, Cj, Liu, Naiming, Luzi, Lorenzo, Mason, Blake, Baraniuk, Richard G.

“…Deep neural networks have become essential for numerous applications due to their strong empirical performance such as vision, RL, and classification…”
Get full text

Wearing A Mask: Compressed Representations of Variable-Length Sequences Using Recurrent Neural Tangent Kernels by Alemohammad, Sina, Babaei, Hossein, Balestriero, Randall, Cheung, Matt Y., Humayun, Ahmed Imtiaz, LeJeune, Daniel, Liu, Naiming, Luzi, Lorenzo, Tan, Jasper, Wang, Zichao, Baraniuk, Richard G.

“…High dimensionality poses many challenges to the use of data, from visualization and interpretation, to prediction and storage for historical preservation…”
Get full text

Frozen Overparameterization: A Double Descent Perspective on Transfer Learning of Deep Neural Networks by Dar, Yehuda, Luzi, Lorenzo, Baraniuk, Richard G

“…We study the generalization behavior of transfer learning of deep neural networks (DNNs). We adopt the overparameterization perspective -- featuring…”
Get full text

Using Skew to Assess the Quality of GAN-generated Image Features by Luzi, Lorenzo, Jenne, Helen, Murray, Ryan, Marrero, Carlos Ortiz

“…The rapid advancement of Generative Adversarial Networks (GANs) necessitates the need to robustly evaluate these models. Among the established evaluation…”
Get full text

Double Descent and Other Interpolation Phenomena in GANs by Luzi, Lorenzo, Dar, Yehuda, Baraniuk, Richard

“…We study overparameterization in generative adversarial networks (GANs) that can interpolate the training data. We show that overparameterization can improve…”
Get full text

Improving Fairness and Mitigating MADness in Generative Models by Mayer, Paul, Luzi, Lorenzo, Siahkoohi, Ali, Johnson, Don H, Baraniuk, Richard G

“…Generative models unfairly penalize data belonging to minority classes, suffer from model autophagy disorder (MADness), and learn biased estimates of the…”
Get full text

Ensembles of Generative Adversarial Networks for Disconnected Data by Luzi, Lorenzo, Balestriero, Randall, Baraniuk, Richard G

“…Most current computer vision datasets are composed of disconnected sets, such as images from different classes. We prove that distributions of this type of…”
Get full text

Boomerang: Local sampling on image manifolds using diffusion models by Luzi, Lorenzo, Mayer, Paul M, Casco-Rodriguez, Josue, Siahkoohi, Ali, Baraniuk, Richard G

“…The inference stage of diffusion models can be seen as running a reverse-time diffusion stochastic differential equation, where samples from a Gaussian latent…”
Get full text

Self-Consuming Generative Models Go MAD by Alemohammad, Sina, Casco-Rodriguez, Josue, Luzi, Lorenzo, Humayun, Ahmed Imtiaz, Babaei, Hossein, LeJeune, Daniel, Siahkoohi, Ali, Baraniuk, Richard G

“…Seismic advances in generative AI algorithms for imagery, text, and other data types has led to the temptation to use synthetic data to train next-generation…”
Get full text

TITAN: Bringing The Deep Image Prior to Implicit Representations by Luzi, Lorenzo, LeJeune, Daniel, Siahkoohi, Ali, Alemohammad, Sina, Saragadam, Vishwanath, Babaei, Hossein, Liu, Naiming, Wang, Zichao, Baraniuk, Richard G

“…We study the interpolation capabilities of implicit neural representations (INRs) of images. In principle, INRs promise a number of advantages, such as…”
Get full text

Subspace Fitting Meets Regression: The Effects of Supervision and Orthonormality Constraints on Double Descent of Generalization Errors by Dar, Yehuda, Mayer, Paul, Luzi, Lorenzo, Baraniuk, Richard G

“…We study the linear subspace fitting problem in the overparameterized setting, where the estimated subspace can perfectly interpolate the training examples…”
Get full text

Evaluating generative networks using Gaussian mixtures of image features by Luzi, Lorenzo, Marrero, Carlos Ortiz, Wynar, Nile, Baraniuk, Richard G, Henry, Michael J

“…We develop a measure for evaluating the performance of generative networks given two sets of images. A popular performance measure currently used to do this is…”
Get full text

NFT-K: Non-Fungible Tangent Kernels by Alemohammad, Sina, Babaei, Hossein, Barberan, CJ, Liu, Naiming, Luzi, Lorenzo, Mason, Blake, Baraniuk, Richard G

“…Deep neural networks have become essential for numerous applications due to their strong empirical performance such as vision, RL, and classification…”
Get full text