# Are Synthetic Corruptions A Reliable Proxy For Real-World Corruptions?

Source: [Are Synthetic Corruptions A Reliable Proxy For Real-World Corruptions?](<https://www.sota2.com/research/paper/68fa787b1e209131cea78a0c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T04:39:15.781Z

Authors: Shashank Agnihotri, David Schader, Nico Sharei, Mehmet Ege Ka\\c{c}ar, Margret Keuper

Published: 2025-05-07T22:19:55.000Z

## Abstract

Deep learning (DL) models are widely used in real-world applications but remain vulnerable to distribution shifts, especially due to weather and lighting changes. Collecting diverse real-world data for testing the robustness of DL models is resource-intensive, making synthetic corruptions an attractive alternative for robustness testing. However, are synthetic corruptions a reliable proxy for real-world corruptions? To answer this, we conduct the largest benchmarking study on semantic segmentation models, comparing performance on real-world corruptions and synthetic corruptions datasets. Our results reveal a strong correlation in mean performance, supporting the use of synthetic corruptions for robustness evaluation. We further analyze corruption-specific correlations, providing key insights to understand when synthetic corruptions succeed in representing real-world corruptions. Open-source Code: https://github.com/shashankskagnihotri/benchmarking\_robustness/tree/segmentation\_david/semantic\_segmentation

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2505.04835>)

[PDF](<https://arxiv.org/pdf/2505.04835.pdf>)