This is a challenging inference task given the need to reason beyond the local appearance of hands. The lack of training annotations indicating which object or parts of an object the hand is in contact with further complicates the task.

The Information Bottleneck (IB) objective uses information theory to formulate a task-performance versus robustness trade-off.

MLPs with"noencoding" struggle tofit high frequencysegments (see appendix for train details). Our workenables MLP networks toadaptivelyfitavarying spectrum offine details that previous methods struggle to capture in a single shot, without involved tuning of parameters or domain specific preprocessing.

We study the generalisation properties of majority voting on finite ensembles of classifiers, proving margin-based generalisation bounds via the PAC-Bayes theory. These provide state-of-the-art guarantees on a number of classification tasks. Our central results leverage the Dirichlet posteriors studied recently by Zantedeschi et al. (2021) for training voting classifiers; in contrast to that work our bounds apply to non-randomised votes via the use of margins. Our contributions add perspective to the debate on the "margins theory" proposed by Schapire et al. (1998) for the generalisation of ensemble classifiers.

The MAB paradigm provides a succinct abstraction of the quintessential exploration vs. exploitation trade-offs inherent in many sequential decision making problems.