English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The top-k classification accuracy is one of the core metrics in machinelearning. Here, k is conventionally a positive integer, such as 1 or 5, leadingto top-1 or top-5 training objectives. In this work, we relax this assumptionand optimize the model for multiple k simultaneously instead of using a singlek. Leveraging recent advances in differentiable sorting and ranking, we proposea differentiable top-k cross-entropy classification loss. This allows trainingthe network while not only considering the top-1 prediction, but also, e.g.,the top-2 and top-5 predictions. We evaluate the proposed loss function forfine-tuning on state-of-the-art architectures, as well as for training fromscratch. We find that relaxing k does not only produce better top-5 accuracies,but also leads to top-1 accuracy improvements. When fine-tuning publiclyavailable ImageNet models, we achieve a new state-of-the-art for these models.

Differentiable Top-k Classification Learning