philipperemy/keras-sde-net

SDE Net (Keras)

This repo contains the code for the paper:

Lingkai Kong, Jimeng Sun and Chao Zhang, SDE-Net: Equipping Deep Neural Network with Uncertainty Estimates, ICML2020.

Package installation

From PyPI

pip install sdenet

From the sources

git clone https://github.com/philipperemy/keras-sde-net.git && cd keras-sde-net
virtualenv -p python3 venv && source venv/bin/activate # optional but recommended.
pip install -r requirements.txt && pip install -e . # install the package.

Import the models

from sdenet import SDENet
from sdenet import ResidualNet

Training & Evaluation

Supported datasets are: MNIST, SVHN, CIFAR10, CIFAR100. Supported models are RESNET and SDENET.

Look at the bash run scripts at the root of the repository to get started for training and evaluation.

Comparison between official Pytorch implementation and Keras

This comparison is just the result of one run. No runs were handpicked. Overall it's very similar.

Except probably SDENET on SVHN (95% vs 94%).

Pytorch

MNIST RESNET
_________________________________

Final Accuracy: 9945/10000 (99.45%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            88.783%
AUROC:                     95.939%
Detection acc:             92.169%
AUPR In:                   86.441%
AUPR Out:                  98.434%

calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            89.791%
AUROC:                     97.510%
Detection acc:             93.041%
AUPR In:                   99.985%
AUPR Out:                  34.000%


MNIST SDENET
_________________________________

Final Accuracy: 9927/10000 (99.27%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            99.372%
AUROC:                     99.804%
Detection acc:             98.692%
AUPR In:                   99.483%
AUPR Out:                  99.887%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            92.544%
AUROC:                     97.525%
Detection acc:             94.485%
AUPR In:                   99.979%
AUPR Out:                  41.739%


SVHN RESNET
_________________________________

Final Accuracy: 24609/25856 (95.18%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            66.552%
AUROC:                     94.421%
Detection acc:             90.136%
AUPR In:                   97.639%
AUPR Out:                  84.998%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            64.376%
AUROC:                     90.458%
Detection acc:             85.371%
AUPR In:                   99.301%
AUPR Out:                  44.899%


SVHN SDENET
_________________________________

Final Accuracy: 24588/25856 (95.10%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            65.215%
AUROC:                     94.308%
Detection acc:             89.746%
AUPR In:                   97.694%
AUPR Out:                  84.017%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            67.831%
AUROC:                     91.267%
Detection acc:             86.501%
AUPR In:                   99.270%
AUPR Out:                  48.871%

Keras

MNIST RESNET
_________________________________

 Final Accuracy: 9944/10000 (99.44%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            93.162%
AUROC:                     97.946%
Detection acc:             94.250%
AUPR In:                   94.842%
AUPR Out:                  99.215%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            96.997%
AUROC:                     98.863%
Detection acc:             96.697%
AUPR In:                   99.994%
AUPR Out:                  26.744%

MNIST SDENET
_________________________________

Final Accuracy: 9934/10000 (99.34%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            98.425%
AUROC:                     99.567%
Detection acc:             97.804%
AUPR In:                   98.613%
AUPR Out:                  99.872%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            95.515%
AUROC:                     98.763%
Detection acc:             95.825%
AUPR In:                   99.992%
AUPR Out:                  32.524%

SVHN RESNET
_________________________________

 Final Accuracy: 24487/25856 (94.71%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            56.648%
AUROC:                     93.602%
Detection acc:             87.504%
AUPR In:                   97.627%
AUPR Out:                  81.664%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            63.765%
AUROC:                     91.843%
Detection acc:             85.721%
AUPR In:                   99.386%
AUPR Out:                  46.231%

SVHN SDENET
_________________________________

 Final Accuracy: 24339/25856 (94.13%)

generate log  from out-of-distribution data
calculate metrics for OOD
OOD  Performance of Baseline detector
TNR at TPR 95%:            64.491%
AUROC:                     94.358%
Detection acc:             88.711%
AUPR In:                   97.776%
AUPR Out:                  87.517%
calculate metrics for mis
mis  Performance of Baseline detector
TNR at TPR 95%:            60.160%
AUROC:                     88.955%
Detection acc:             85.735%
AUPR In:                   99.165%
AUPR Out:                  45.268%

philipperemy / keras-sde-net