# Image Classification on ImageNet-200 (test) Source: [Image Classification on ImageNet-200 (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:07:19.191Z Evaluation of Image Classification performance on the test split of the ImageNet-200 subset, measured using metrics such as Top-1 Error Rate. 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-200 (test) | | Metric | Top-1 Error Rate | | Unit | PERCENT | | Direction | Lower is better | | Results loaded | 28 of 28 | | 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 - [Top-1 Error Rate](): Lower is better; unit: PERCENT. ## Recorded results | Method | Top-1 Error Rate | Result date | Report paper | | --- | --- | --- | --- | | ResNeXt-101 32x8d + WSL Pretraining | 4.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNeXt-101 32x4d + SE | 5.9 | Not recorded | 68fa6a751e209131cea28405 | | ResNeXt-101 32x8d + DeepAugment + AugMix | 6.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNeXt-101 32x8d | 6.2 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-152 + SE | 6.6 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-101 + SE | 6.7 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-152 | 6.8 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-152 | 6.8 | Not recorded | 68fa6a751e209131cea28405 | | ResNeXt-101 32x4d | 6.8 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + ImageNet-21K Pretraining | 7 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + CBAM | 7 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + ImageNet-21K Pretraining | 7 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + CBAM | 7 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + AugMix | 7.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + AugMix | 7.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-101 | 7.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + DeepAugment | 7.5 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + DeepAugment | 7.5 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 | 7.9 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 | 7.9 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + DeepAugment + AugMix | 8 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + DeepAugment + AugMix | 8 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + Speckle Noise | 8.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + Speckle Noise | 8.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + Style Transfer Augmentation | 8.9 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + Style Transfer | 8.9 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + l\_inf Adversarial Training | 25.1 | Not recorded | 68fa6a751e209131cea28405 | | ResNet-50 + L\_inf Adversarial Training | 25.1 | Not recorded | 68fa6a751e209131cea28405 | ### ResNeXt-101 32x8d + WSL Pretraining Top-1 Error Rate: 4.1. Result ID: 698e36c9ffc3901de09bb4ae. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNeXt-101 32x8d","Pre-training":"WSL Pretraining (1000x data)"}. This record does not assert independent reproduction. ### ResNeXt-101 32x4d + SE Top-1 Error Rate: 5.9. Result ID: 698e36c9ffc3901de09bb4ac. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNeXt-101 32x4d","Self-Attention":"SE"}. This record does not assert independent reproduction. ### ResNeXt-101 32x8d + DeepAugment + AugMix Top-1 Error Rate: 6.1. Result ID: 698e36c9ffc3901de09bb4af. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNeXt-101 32x8d","Data Augmentation":"DeepAugment + AugMix"}. This record does not assert independent reproduction. ### ResNeXt-101 32x8d Top-1 Error Rate: 6.2. Result ID: 698e36c9ffc3901de09bb4ad. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNeXt-101 32x8d"}. This record does not assert independent reproduction. ### ResNet-152 + SE Top-1 Error Rate: 6.6. Result ID: 698e36c9ffc3901de09bb4aa. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-152","Self-Attention":"SE"}. This record does not assert independent reproduction. ### ResNet-101 + SE Top-1 Error Rate: 6.7. Result ID: 698e36c9ffc3901de09bb4a8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-101","Self-Attention":"SE"}. This record does not assert independent reproduction. ### ResNet-152 Top-1 Error Rate: 6.8. Result ID: 698e36c9ffc3901de09bb493. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-152"}. This record does not assert independent reproduction. ### ResNet-152 Top-1 Error Rate: 6.8. Result ID: 698e36c9ffc3901de09bb4a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-152"}. This record does not assert independent reproduction. ### ResNeXt-101 32x4d Top-1 Error Rate: 6.8. Result ID: 698e36c9ffc3901de09bb4ab. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNeXt-101 32x4d"}. This record does not assert independent reproduction. ### ResNet-50 + ImageNet-21K Pretraining Top-1 Error Rate: 7. Result ID: 698e36c9ffc3901de09bb48b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-training":"ImageNet-21K"}. This record does not assert independent reproduction. ### ResNet-50 + CBAM Top-1 Error Rate: 7. Result ID: 698e36c9ffc3901de09bb48c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Attention Mechanism":"CBAM (Self-Attention)"}. This record does not assert independent reproduction. ### ResNet-50 + ImageNet-21K Pretraining Top-1 Error Rate: 7. Result ID: 698e36c9ffc3901de09bb49f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-training":"ImageNet-21K (10x data)"}. This record does not assert independent reproduction. ### ResNet-50 + CBAM Top-1 Error Rate: 7. Result ID: 698e36c9ffc3901de09bb4a0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Self-Attention":"CBAM"}. This record does not assert independent reproduction. ### ResNet-50 + AugMix Top-1 Error Rate: 7.1. Result ID: 698e36c9ffc3901de09bb490. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"AugMix"}. This record does not assert independent reproduction. ### ResNet-50 + AugMix Top-1 Error Rate: 7.1. Result ID: 698e36c9ffc3901de09bb4a4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"AugMix"}. This record does not assert independent reproduction. ### ResNet-101 Top-1 Error Rate: 7.1. Result ID: 698e36c9ffc3901de09bb4a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-101"}. This record does not assert independent reproduction. ### ResNet-50 + DeepAugment Top-1 Error Rate: 7.5. Result ID: 698e36c9ffc3901de09bb491. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"DeepAugment"}. This record does not assert independent reproduction. ### ResNet-50 + DeepAugment Top-1 Error Rate: 7.5. Result ID: 698e36c9ffc3901de09bb4a5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"DeepAugment"}. This record does not assert independent reproduction. ### ResNet-50 Top-1 Error Rate: 7.9. Result ID: 698e36c9ffc3901de09bb48a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction. ### ResNet-50 Top-1 Error Rate: 7.9. Result ID: 698e36c9ffc3901de09bb49e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction. ### ResNet-50 + DeepAugment + AugMix Top-1 Error Rate: 8. Result ID: 698e36c9ffc3901de09bb492. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"DeepAugment + AugMix"}. This record does not assert independent reproduction. ### ResNet-50 + DeepAugment + AugMix Top-1 Error Rate: 8. Result ID: 698e36c9ffc3901de09bb4a6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"DeepAugment + AugMix"}. This record does not assert independent reproduction. ### ResNet-50 + Speckle Noise Top-1 Error Rate: 8.1. Result ID: 698e36c9ffc3901de09bb48e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"Speckle Noise"}. This record does not assert independent reproduction. ### ResNet-50 + Speckle Noise Top-1 Error Rate: 8.1. Result ID: 698e36c9ffc3901de09bb4a2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"Speckle Noise"}. This record does not assert independent reproduction. ### ResNet-50 + Style Transfer Augmentation Top-1 Error Rate: 8.9. Result ID: 698e36c9ffc3901de09bb48f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"Style Transfer Augmentation"}. This record does not assert independent reproduction. ### ResNet-50 + Style Transfer Top-1 Error Rate: 8.9. Result ID: 698e36c9ffc3901de09bb4a3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Data Augmentation":"Style Transfer"}. This record does not assert independent reproduction. ### ResNet-50 + l\_inf Adversarial Training Top-1 Error Rate: 25.1. Result ID: 698e36c9ffc3901de09bb48d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Training Strategy":"l\_inf Adversarial Training"}. This record does not assert independent reproduction. ### ResNet-50 + L\_inf Adversarial Training Top-1 Error Rate: 25.1. Result ID: 698e36c9ffc3901de09bb4a1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Training Strategy":"L\_inf Adversarial Training"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.