AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Intra Order-preserving Functions for Calibration of Multi-Class Neural Networks

Amir Rahimi, Amirreza Shaban, Ching-An Cheng, Richard Hartley, Byron Boots

NeurIPS 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Confidence calibration objectives · 21 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
learn calibration function within intra order-preserving family (Diag) instead of temperature scalingpost-hoc calibration · 14 dataset-model pairsaverage relative ECE: 0.42 -> 0.27Improved
use order-invariant order-preserving function (OI) instead of temperature scalingpost-hoc calibration · 14 dataset-model pairsaverage relative ECE: 0.42 -> 0.33Improved
add order-preserving constraint (OP) vs unconstrained MLP calibrator (ablation)MLP calibration function · 14 dataset-model pairsaverage relative ECE: 0.66 -> 0.41Improved
unconstrained MLP calibrator hurts accuracy (ablation)MLP calibration function · ImageNet ResNet-152accuracy change: -33.4%Hurt
add order-invariant sharing (OI) to general order-preserving (OP) (ablation)intra order-preserving calibrator · 14 dataset-model pairsaverage relative ECE: 0.41 -> 0.33Improved
add diagonal parameter sharing (Diag) to OI (ablation)intra order-preserving calibrator · 14 dataset-model pairsaverage relative ECE: 0.33 -> 0.27Improved
Diag vs Dirichlet calibration on classwise-ECEpost-hoc calibration · average relative Classwise-ECErelative Classwise-ECE: 0.704 -> 0.729Hurt
Diag vs temperature scaling on NLLpost-hoc calibration · average relative NLLrelative NLL: 0.766 -> 0.749Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy