Distribution Shifts at Scale: Out-of-distribution Detection in Earth Observation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ekim, Burak, Tadesse, Girmaw Abebe, Robinson, Caleb, Hacheme, Gilles, Schmitt, Michael, Dodhia, Rahul, Ferres, Juan M. Lavista
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913784139874304
author Ekim, Burak
Tadesse, Girmaw Abebe
Robinson, Caleb
Hacheme, Gilles
Schmitt, Michael
Dodhia, Rahul
Ferres, Juan M. Lavista
author_facet Ekim, Burak
Tadesse, Girmaw Abebe
Robinson, Caleb
Hacheme, Gilles
Schmitt, Michael
Dodhia, Rahul
Ferres, Juan M. Lavista
contents Training robust deep learning models is crucial in Earth Observation, where globally deployed models often face distribution shifts that degrade performance, especially in low-data regions. Out-of-distribution (OOD) detection addresses this by identifying inputs that deviate from in-distribution (ID) data. However, existing methods either assume access to OOD data or compromise primary task performance, limiting real-world use. We introduce TARDIS, a post-hoc OOD detection method designed for scalable geospatial deployment. Our core innovation lies in generating surrogate distribution labels by leveraging ID data within the feature space. TARDIS takes a pre-trained model, ID data, and data from an unknown distribution (WILD), separates WILD into surrogate ID and OOD labels based on internal activations, and trains a binary classifier to detect distribution shifts. We validate on EuroSAT and xBD across 17 setups covering covariate and semantic shifts, showing near-upper-bound surrogate labeling performance in 13 cases and matching the performance of top post-hoc activation- and scoring-based methods. Finally, deploying TARDIS on Fields of the World reveals actionable insights into pre-trained model behavior at scale. The code is available at \href{https://github.com/microsoft/geospatial-ood-detection}{https://github.com/microsoft/geospatial-ood-detection}
format Preprint
id arxiv_https___arxiv_org_abs_2412_13394
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Distribution Shifts at Scale: Out-of-distribution Detection in Earth Observation
Ekim, Burak
Tadesse, Girmaw Abebe
Robinson, Caleb
Hacheme, Gilles
Schmitt, Michael
Dodhia, Rahul
Ferres, Juan M. Lavista
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Training robust deep learning models is crucial in Earth Observation, where globally deployed models often face distribution shifts that degrade performance, especially in low-data regions. Out-of-distribution (OOD) detection addresses this by identifying inputs that deviate from in-distribution (ID) data. However, existing methods either assume access to OOD data or compromise primary task performance, limiting real-world use. We introduce TARDIS, a post-hoc OOD detection method designed for scalable geospatial deployment. Our core innovation lies in generating surrogate distribution labels by leveraging ID data within the feature space. TARDIS takes a pre-trained model, ID data, and data from an unknown distribution (WILD), separates WILD into surrogate ID and OOD labels based on internal activations, and trains a binary classifier to detect distribution shifts. We validate on EuroSAT and xBD across 17 setups covering covariate and semantic shifts, showing near-upper-bound surrogate labeling performance in 13 cases and matching the performance of top post-hoc activation- and scoring-based methods. Finally, deploying TARDIS on Fields of the World reveals actionable insights into pre-trained model behavior at scale. The code is available at \href{https://github.com/microsoft/geospatial-ood-detection}{https://github.com/microsoft/geospatial-ood-detection}
title Distribution Shifts at Scale: Out-of-distribution Detection in Earth Observation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2412.13394