# Image Classification on ImageNet (test) (Corruption Accuracy and mCE) Source: [Image Classification on ImageNet (test) (Corruption Accuracy and mCE)]() Collection index: [llms.txt]() Data updated: 2026-02-27T08:00:05.783Z Evaluation of Image Classification performance on the test split of ImageNet, specifically reporting metrics related to model robustness against common corruptions (Clean Acc, Common Corruption Acc, and mean Corruption Error (mCE)). 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 (test) | | Metric | Clean Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 10 of 10 | | 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 - [Clean Accuracy](): Higher is better; unit: PERCENT. - [Common Corruption Accuracy](): Higher is better; unit: PERCENT. - [Mean Corruption Error](): Lower is better; unit: PERCENT. - [Corruption Acc L1](): Higher is better; unit: PERCENT. - [Corruption Acc L2](): Higher is better; unit: PERCENT. - [Corruption Acc L3](): Higher is better; unit: PERCENT. - [Corruption Acc L4](): Higher is better; unit: PERCENT. - [Corruption Acc L5](): Higher is better; unit: PERCENT. - [Corruption Acc L6](): Higher is better; unit: PERCENT. ## Recorded results | Method | Clean Accuracy | Result date | Report paper | | --- | --- | --- | --- | | AugMix | 77.5 | Not recorded | 68fa6d351e209131cea386e2 | | PRIME | 77 | Not recorded | 68fa6d351e209131cea386e2 | | DeepAugment (DA) | 76.7 | Not recorded | 68fa6d351e209131cea386e2 | | Standard | 76.1 | Not recorded | 68fa6d351e209131cea386e2 | | DA+AugMix | 75.8 | Not recorded | 68fa6d351e209131cea386e2 | | DA+PRIME | 75.5 | Not recorded | 68fa6d351e209131cea386e2 | ### AugMix Clean Accuracy: 77.5. Result ID: 698f83ed3a6c314849f9913a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Source Model Reference":"\[9\]"}. This record does not assert independent reproduction. ### PRIME Clean Accuracy: 77. Result ID: 698f83ed3a6c314849f9913c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","JSD Consistency Loss":false}. This record does not assert independent reproduction. ### DeepAugment (DA) Clean Accuracy: 76.7. Result ID: 698f83ed3a6c314849f9913b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Source Model Reference":"\[9\]"}. This record does not assert independent reproduction. ### Standard Clean Accuracy: 76.1. Result ID: 698f83ed3a6c314849f99139. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Source Model Reference":"\[9\]"}. This record does not assert independent reproduction. ### DA+AugMix Clean Accuracy: 75.8. Result ID: 698f83ed3a6c314849f9913d. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction. ### DA+PRIME Clean Accuracy: 75.5. Result ID: 698f83ed3a6c314849f9913e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","JSD Consistency Loss":false}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.