# Image Classification on Cars, EuroSAT, RESISC45, GTSRB (Corrupted Test Average) Source: [Image Classification on Cars, EuroSAT, RESISC45, GTSRB (Corrupted Test Average)]() Collection index: [llms.txt]() Data updated: 2026-04-28T08:20:38.360Z Aggregate evaluation of Image Classification performance across the test splits of Cars, EuroSAT, RESISC45, and GTSRB, specifically measuring robustness against various common image corruptions (e.g., blur, noise, compression). 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) | Cars, EuroSAT, RESISC45, GTSRB Average - Corrupted (test) | | Metric | Accuracy (Motion Blur) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 12 of 12 | | 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 (Motion Blur)](): Higher is better; unit: PERCENT. - [Accuracy (Impulse Noise)](): Higher is better; unit: PERCENT. - [Accuracy (Gaussian Noise)](): Higher is better; unit: PERCENT. - [Accuracy (Pixelate)](): Higher is better; unit: PERCENT. - [Accuracy (Spatter)](): Higher is better; unit: PERCENT. - [Accuracy (Contrast)](): Higher is better; unit: PERCENT. - [Accuracy (JPEG Compression)](): Higher is better; unit: PERCENT. - [Average Accuracy (Corrupted Test)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (Motion Blur) | Result date | Report paper | | --- | --- | --- | --- | | DOGE AM | 85 | Not recorded | 68fa7a631e209131cea82cfa | | RegMean++ | 81.8 | Not recorded | 68fa7a631e209131cea82cfa | | Layer-wise AdaMerging | 81.3 | Not recorded | 68fa7a631e209131cea82cfa | | RegMean | 79.7 | Not recorded | 68fa7a631e209131cea82cfa | | TSV-M | 78.9 | Not recorded | 68fa7a631e209131cea82cfa | | DOGE TA | 77.3 | Not recorded | 68fa7a631e209131cea82cfa | | Iso-C | 75.9 | Not recorded | 68fa7a631e209131cea82cfa | | Iso-CTS | 75.7 | Not recorded | 68fa7a631e209131cea82cfa | | Fisher Merging | 67 | Not recorded | 68fa7a631e209131cea82cfa | | Task Arithmetic | 65.9 | Not recorded | 68fa7a631e209131cea82cfa | | Model Soups | 64.6 | Not recorded | 68fa7a631e209131cea82cfa | | TIES-Merging | 63.2 | Not recorded | 68fa7a631e209131cea82cfa | ### DOGE AM Accuracy (Motion Blur): 85. Result ID: 69f06cb147f34ce924d63b85. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Test-Time Adaptation"}. This record does not assert independent reproduction. ### RegMean++ Accuracy (Motion Blur): 81.8. Result ID: 69f06cb147f34ce924d63b83. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Training-Free"}. This record does not assert independent reproduction. ### Layer-wise AdaMerging Accuracy (Motion Blur): 81.3. Result ID: 69f06cb147f34ce924d63b84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Test-Time Adaptation"}. This record does not assert independent reproduction. ### RegMean Accuracy (Motion Blur): 79.7. Result ID: 69f06cb147f34ce924d63b82. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Training-Free"}. This record does not assert independent reproduction. ### TSV-M Accuracy (Motion Blur): 78.9. Result ID: 69f06cb147f34ce924d63b7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### DOGE TA Accuracy (Motion Blur): 77.3. Result ID: 69f06cb147f34ce924d63b7e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### Iso-C Accuracy (Motion Blur): 75.9. Result ID: 69f06cb147f34ce924d63b7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### Iso-CTS Accuracy (Motion Blur): 75.7. Result ID: 69f06cb147f34ce924d63b80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### Fisher Merging Accuracy (Motion Blur): 67. Result ID: 69f06cb147f34ce924d63b81. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Training-Free"}. This record does not assert independent reproduction. ### Task Arithmetic Accuracy (Motion Blur): 65.9. Result ID: 69f06cb147f34ce924d63b7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### Model Soups Accuracy (Motion Blur): 64.6. Result ID: 69f06cb147f34ce924d63b7a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. ### TIES-Merging Accuracy (Motion Blur): 63.2. Result ID: 69f06cb147f34ce924d63b7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ViT-B/32","Method Category":"Data-Free"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.