← Search

Ruairidh M. Battleday

2 accepted papers

2023

On the informativeness of supervision signals

UAI 2023poster

Supervised learning typically focuses on learning transferable representations from training examples annotated by humans. While rich annotations (like soft labels) carry more information than sparse annotations (like hard labels), they are also more expensive to collect. For example, while hard lab…

Cited by 17SourcePDFScholar
2019

Human Uncertainty Makes Classification More Robust

ICCV 2019poster

The classification performance of deep neural networks has begun to asymptote at near-perfect levels. However, their ability to generalize outside the training set and their robustness to adversarial attacks have not. In this paper, we make progress on this problem by training with full label distri…

Cited by 388PDFcodeScholar