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 great performances of deep learning are undeniable, with impressiveresults over a wide range of tasks. However, the output confidence of thesemodels is usually not well-calibrated, which can be an issue for applicationswhere confidence on the decisions is central to providing trust and reliability(e.g., autonomous driving or medical diagnosis). For models using softmax atthe last layer, Temperature Scaling (TS) is a state-of-the-art calibrationmethod, with low time and memory complexity as well as demonstratedeffectiveness. TS relies on a T parameter to rescale and calibrate values ofthe softmax layer, whose parameter value is computed from a labelled dataset.We are proposing an Unsupervised Temperature Scaling (UTS) approach, which doesnot depend on labelled samples to calibrate the model, which allows, forexample, the use of a part of a test samples to calibrate the pre-trained modelbefore going into inference mode. We provide theoretical justifications for UTSand assess its effectiveness on a wide range of deep models and datasets. Wealso demonstrate calibration results of UTS on skin lesion detection, a problemwhere a well-calibrated output can play an important role for accuratedecision-making.

Unsupervised Temperature Scaling: An Unsupervised Post-Processing Calibration Method of Deep Networks