# Image Classification on ImageNet (Corruption Robustness Metrics) Source: [Image Classification on ImageNet (Corruption Robustness Metrics)]() Collection index: [llms.txt]() Data updated: 2026-02-27T08:00:05.575Z Evaluation of Image Classification performance on the standard ImageNet dataset, focusing on robustness metrics derived from common image corruptions (e.g., Hue, Saturation, Brightness, Contrast) and specific perturbation metrics (l-infinity norms). 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. | Scope | Recorded value | | --- | --- | | Task | Image Classification | | Dataset (including any stated split) | ImageNet | | Metric | Accuracy (Clean) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 4 of 4 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy (Clean)](): Higher is better; unit: PERCENT. - [Accuracy (Hue Corruption)](): Higher is better; unit: PERCENT. - [Accuracy (Saturation Corruption)](): Higher is better; unit: PERCENT. - [Accuracy (Rotation)](): Higher is better; unit: PERCENT. - [Accuracy (Brightness Corruption)](): Higher is better; unit: PERCENT. - [Accuracy (Contrast Corruption)](): Higher is better; unit: PERCENT. - [Accuracy (L-inf, 20-step)](): Higher is better; unit: PERCENT. - [Accuracy (Auto L-inf)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (Clean) | Result date | Report paper | | --- | --- | --- | --- | | Normal | 0 | Not recorded | 68fa6ecc1e209131cea42343 | | Madry | 0 | Not recorded | 68fa6ecc1e209131cea42343 | | Fast-AT | 0 | Not recorded | 68fa6ecc1e209131cea42343 | | GAT-f | 0 | Not recorded | 68fa6ecc1e209131cea42343 | ### Normal Accuracy (Clean): 0. Result ID: 69a14ef535dfcabb00e62742. [Reporting paper]() Recorded evaluation configuration: {"Training":"Normal†"}. This record does not assert independent reproduction. ### Madry Accuracy (Clean): 0. Result ID: 69a14ef535dfcabb00e62743. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"Madry†∞"}. This record does not assert independent reproduction. ### Fast-AT Accuracy (Clean): 0. Result ID: 69a14ef535dfcabb00e62744. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"Fast-AT†∞"}. This record does not assert independent reproduction. ### GAT-f Accuracy (Clean): 0. Result ID: 69a14ef535dfcabb00e62745. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Training":"GAT-f†"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.