SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2202.03299 · ICML · 2022

Training OOD Detectors in their Natural Habitats

Robert Nowak, Yixuan Li, Julian Katz-Samuels, Julia Nakhleh

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 14 functions out of this paper's own repositories and ran 8 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
jkatzsam/woods_ood canonical 8 of 14
FunctionStatusWhere it lives
ODIN Ran jkatzsam/woods_ood/utils/score_calculation.py
code served (permissive licence) · get_code("17bbd961aaf86569")
compute_auroc Ran jkatzsam/woods_ood/CIFAR/plot_results.py
code served (permissive licence) · get_code("581a8583346c6d6e")
fpr_and_fdr_at_recall Ran jkatzsam/woods_ood/utils/display_results.py
code served (permissive licence) · get_code("7242da839b9822be")
get_Mahalanobis_score Ran jkatzsam/woods_ood/utils/score_calculation.py
code served (permissive licence) · get_code("24f5b40495bb542e")
get_measures Ran jkatzsam/woods_ood/utils/display_results.py
code served (permissive licence) · get_code("b9b6c36f7a3e7e67")
get_ood_scores_odin Ran jkatzsam/woods_ood/utils/score_calculation.py
code served (permissive licence) · get_code("3ad01de2f86dccba")
stable_cumsum Ran jkatzsam/woods_ood/utils/display_results.py
code served (permissive licence) · get_code("d4acb3120a027622")
test_fnr_using_valid_threshold Ran jkatzsam/woods_ood/CIFAR/plot_results.py
code served (permissive licence) · get_code("2b5f3b49bd6f6768")
calib_err Not yet run jkatzsam/woods_ood/utils/calibration_tools.py
code served (permissive licence) · get_code("509aa629e46cf950")
load_CIFAR Not yet run jkatzsam/woods_ood/CIFAR/make_datasets.py
code served (permissive licence) · get_code("2200861da0136416")
load_results Not yet run jkatzsam/woods_ood/CIFAR/plot_results.py
code served (permissive licence) · get_code("b2dd765101838953")
soft_f1 Not yet run jkatzsam/woods_ood/utils/calibration_tools.py
code served (permissive licence) · get_code("0a958e667c8c24c8")
tune_temp Not yet run jkatzsam/woods_ood/utils/calibration_tools.py
code served (permissive licence) · get_code("3c09c6a69241b742")
unpickle Not yet run jkatzsam/woods_ood/utils/imagenet_rc_loader.py
code served (permissive licence) · get_code("4ebc0e09cb353eac")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Out-of-distribution (OOD) detection is important for machine learning models deployed in the wild. Recent methods use auxiliary outlier data to regularize the model for improved OOD detection. However, these approaches make a strong distributional assumption that the auxiliary outlier data is completely separable from the in-distribution (ID) data. In this paper, we propose a novel framework that leverages wild mixture data-that naturally consists of both ID and OOD samples. Such wild data is abundant and arises freely upon deploying a machine learning classifier in their natural habitats. Our key idea is to formulate a constrained optimization problem and to show how to tractably solve it. Our learning objective maximizes the OOD detection rate, subject to constraints on the classification error of ID data and on the OOD error rate of ID examples. We extensively evaluate our approach on common OOD detection tasks and demonstrate superior performance. Code is available at https: //github.com/jkatzsam/woods_ood.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2202.03299")
get_code_for_paper("2202.03299")
have("2202.03299")

Connect an agent — have() is free.