5 papers
Pre-train to Gain: Robust Learning Without Clean Labels
David Szczecina, Nicholas Pellegrino, Paul Fieguth
Training deep networks with noisy labels leads to poor generalization and degraded accuracy due to overfitting to label noise. Existing approaches for learning with noisy labels of…
Effects of Initialization Biases on Deep Neural Network Training Dynamics
Nicholas Pellegrino, David Szczecina, Paul W. Fieguth
Untrained large neural networks, just after random initialization, tend to favour a small subset of classes, assigning high predicted probabilities to these few classes and approxi…
Hard Samples, Bad Labels: Robust Loss Functions That Know When to Back Off
Nicholas Pellegrino, David Szczecina, Paul Fieguth
Incorrectly labelled training data are frustratingly ubiquitous in both benchmark and specially curated datasets. Such mislabelling clearly adversely affects the performance and ge…
Hyperbolic Multimodal Representation Learning for Biological Taxonomies
ZeMing Gong, Chuanqi Tang, Xiaoliang Huo +6
Taxonomic classification in biodiversity research involves organizing biological specimens into structured hierarchies based on evidence, which can come from multiple modalities su…
Particle-Filtering-based Latent Diffusion for Inverse Problems
Amir Nazemi, Mohammad Hadi Sepanj, Nicholas Pellegrino +2
Current strategies for solving image-based inverse problems apply latent diffusion models to perform posterior sampling.However, almost all approaches make no explicit attempt to e…