• Corpus ID: 239998219

A Geometric Perspective towards Neural Calibration via Sensitivity Decomposition

  title={A Geometric Perspective towards Neural Calibration via Sensitivity Decomposition},
  author={Junjiao Tian and Dylan Yung and Yen-Chang Hsu and Zsolt Kira},
It is well known that vision classification models suffer from poor calibration in the face of data distribution shifts. In this paper, we take a geometric approach to this problem. We propose Geometric Sensitivity Decomposition (GSD) which decomposes the norm of a sample feature embedding and the angular similarity to a target classifier into an instance-dependent and an instance-independent component. The instance-dependent component captures the sensitive information about changes in the… 
Exploring Covariate and Concept Shift for Detection and Calibration of Out-of-Distribution Data
This work proposes to characterize the spectrum of OOD data using two types of distribution shifts: covariate shift and concept shift, and proposes a geometricallyinspired method (Geometric ODIN) to improve OOD detection under both shifts with only in-distribution data.


Angular Visual Hardness
Recent convolutional neural networks (CNNs) have led to impressive performance but often suffer from poor calibration. They tend to be overconfident, with the model confidence not always reflecting
Decoupled Networks
  • Weiyang Liu, Z. Liu, +5 authors Le Song
  • Computer Science, Mathematics
    2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
  • 2018
A generic decoupled learning framework which models the intra-class variation and semantic difference independently and directly learn the operator from data is proposed.
Deep Residual Learning for Image Recognition
This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth.
Multi-Class Active Learning by Uncertainty Sampling with Diversity Maximization
This paper proposes a semi-supervised batch mode multi-class active learning algorithm for visual concept recognition that exploits the whole active pool to evaluate the uncertainty of the data, and proposes to make the selected data as diverse as possible.
What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
A Bayesian deep learning framework combining input-dependent aleatoric uncertainty together with epistemic uncertainty is presented, which makes the loss more robust to noisy data, also giving new state-of-the-art results on segmentation and depth regression benchmarks.
On Calibration of Modern Neural Networks
It is discovered that modern neural networks, unlike those from a decade ago, are poorly calibrated, and on most datasets, temperature scaling -- a single-parameter variant of Platt Scaling -- is surprisingly effective at calibrating predictions.
Large-Margin Softmax Loss for Convolutional Neural Networks
A generalized large-margin softmax (L-Softmax) loss which explicitly encourages intra-class compactness and inter-class separability between learned features and which not only can adjust the desired margin but also can avoid overfitting is proposed.
Learning Multiple Layers of Features from Tiny Images
It is shown how to train a multi-layer generative model that learns to extract meaningful features which resemble those found in the human visual cortex, using a novel parallelization algorithm to distribute the work among multiple machines connected on a network.
Improved Training of Wasserstein GANs
This work proposes an alternative to clipping weights: penalize the norm of gradient of the critic with respect to its input, which performs better than standard WGAN and enables stable training of a wide variety of GAN architectures with almost no hyperparameter tuning.
Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
A new theoretical framework is developed casting dropout training in deep neural networks (NNs) as approximate Bayesian inference in deep Gaussian processes, which mitigates the problem of representing uncertainty in deep learning without sacrificing either computational complexity or test accuracy.