# Image Classification on ImageNet Mini Source: [Image Classification on ImageNet Mini]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:07:09.612Z Evaluation of standard Image Classification performance on the ImageNet Mini dataset, typically reporting Top-1 and Top-3 accuracy. 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 Mini | | Metric | Top-1 Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 27 of 27 | | 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 Accuracy](): Higher is better; unit: PERCENT. - [Top-3 Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Top-1 Accuracy | Result date | Report paper | | --- | --- | --- | --- | | CLUENet (base) | 82.44 | Not recorded | 6937ad296cfd488761a4e9c4 | | CLUENet (small) | 81.49 | Not recorded | 6937ad296cfd488761a4e9c4 | | CLUENet (tiny) | 80.51 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientFormer (s2) | 79.75 | Not recorded | 6937ad296cfd488761a4e9c4 | | ShuffleNetv2 (x2.0) | 79.63 | Not recorded | 6937ad296cfd488761a4e9c4 | | FEC (large) | 79.33 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientFormer (s1) | 78.97 | Not recorded | 6937ad296cfd488761a4e9c4 | | CLUENet (micro) | 78.75 | Not recorded | 6937ad296cfd488761a4e9c4 | | ShuffleNetv2 (x1.5) | 78.39 | Not recorded | 6937ad296cfd488761a4e9c4 | | CoC (medium) | 78.39 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientFormer (s0) | 77.92 | Not recorded | 6937ad296cfd488761a4e9c4 | | FEC (base) | 77.81 | Not recorded | 6937ad296cfd488761a4e9c4 | | PVTv2 (b1) | 77.57 | Not recorded | 6937ad296cfd488761a4e9c4 | | ResNet18 | 76.95 | Not recorded | 6937ad296cfd488761a4e9c4 | | FEC (small) | 76.74 | Not recorded | 6937ad296cfd488761a4e9c4 | | CoC (small) | 76.56 | Not recorded | 6937ad296cfd488761a4e9c4 | | PVTv2 (b0) | 75.34 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientViT (m5) | 75.32 | Not recorded | 6937ad296cfd488761a4e9c4 | | ConvNeXtv2 (N) | 75.04 | Not recorded | 6937ad296cfd488761a4e9c4 | | CoC (tiny) | 74.9 | Not recorded | 6937ad296cfd488761a4e9c4 | | ClusterFormer (tiny) | 73.99 | Not recorded | 6937ad296cfd488761a4e9c4 | | ClusterFormer | 73.93 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientViT (m3) | 73.88 | Not recorded | 6937ad296cfd488761a4e9c4 | | EfficientViT (m2) | 73.59 | Not recorded | 6937ad296cfd488761a4e9c4 | | ConvNeXtv2 (F) | 73.14 | Not recorded | 6937ad296cfd488761a4e9c4 | | ConvNeXtv2 (A) | 71.15 | Not recorded | 6937ad296cfd488761a4e9c4 | | ClusterFormer | 70.73 | Not recorded | 6937ad296cfd488761a4e9c4 | ### CLUENet (base) Top-1 Accuracy: 82.44. Result ID: 699883125698665ec4e45408. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"30.20","FLOPs":"6.40","FPS":"679.05±7.12","Memory":"7.7","Type":"Clustering-based"}. This record does not assert independent reproduction. ### CLUENet (small) Top-1 Accuracy: 81.49. Result ID: 699883125698665ec4e45407. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"15.05","FLOPs":"3.16","FPS":"878.60±21.68","Memory":"6.5","Type":"Clustering-based"}. This record does not assert independent reproduction. ### CLUENet (tiny) Top-1 Accuracy: 80.51. Result ID: 699883125698665ec4e45406. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"5.68","FLOPs":"1.30","FPS":"867.76±17.89","Memory":"5.5","Type":"Clustering-based"}. This record does not assert independent reproduction. ### EfficientFormer (s2) Top-1 Accuracy: 79.75. Result ID: 699883125698665ec4e453f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"12.16","FLOPs":"1.27","FPS":"695.41±1.71","Memory":"1.5","Type":"Attention-based"}. This record does not assert independent reproduction. ### ShuffleNetv2 (x2.0) Top-1 Accuracy: 79.63. Result ID: 699883125698665ec4e453f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"6.52","FLOPs":"0.64","FPS":"1205.17±7.59","Memory":"1.3","Type":"Convolution-based"}. This record does not assert independent reproduction. ### FEC (large) Top-1 Accuracy: 79.33. Result ID: 699883125698665ec4e45404. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"29.26","FLOPs":"6.55","FPS":"550.69±2.58","Memory":"7.6","Type":"Clustering-based"}. This record does not assert independent reproduction. ### EfficientFormer (s1) Top-1 Accuracy: 78.97. Result ID: 699883125698665ec4e453f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"5.76","FLOPs":"0.66","FPS":"1128.59±0.75","Memory":"1.5","Type":"Attention-based"}. This record does not assert independent reproduction. ### CLUENet (micro) Top-1 Accuracy: 78.75. Result ID: 699883125698665ec4e45405. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"3.02","FLOPs":"0.65","FPS":"871.89±13.84","Memory":"4.2","Type":"Clustering-based"}. This record does not assert independent reproduction. ### ShuffleNetv2 (x1.5) Top-1 Accuracy: 78.39. Result ID: 699883125698665ec4e453ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"2.58","FLOPs":"0.31","FPS":"1194.72±23.26","Memory":"1.1","Type":"Convolution-based"}. This record does not assert independent reproduction. ### CoC (medium) Top-1 Accuracy: 78.39. Result ID: 699883125698665ec4e453fe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"28.83","FLOPs":"5.96","FPS":"605.67±3.18","Memory":"4.2","Type":"Clustering-based"}. This record does not assert independent reproduction. ### EfficientFormer (s0) Top-1 Accuracy: 77.92. Result ID: 699883125698665ec4e453f6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"3.26","FLOPs":"0.40","FPS":"1166.09±21.46","Memory":"1.5","Type":"Attention-based"}. This record does not assert independent reproduction. ### FEC (base) Top-1 Accuracy: 77.81. Result ID: 699883125698665ec4e45403. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"14.63","FLOPs":"3.37","FPS":"859.63±16.23","Memory":"5.3","Type":"Clustering-based"}. This record does not assert independent reproduction. ### PVTv2 (b1) Top-1 Accuracy: 77.57. Result ID: 699883125698665ec4e453f5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"13.55","FLOPs":"2.06","FPS":"1154.58±39.86","Memory":"2.9","Type":"Attention-based"}. This record does not assert independent reproduction. ### ResNet18 Top-1 Accuracy: 76.95. Result ID: 699883125698665ec4e453ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"14.17","FLOPs":"2.38","FPS":"1150.75±27.15","Memory":"2.7","Type":"Convolution-based"}. This record does not assert independent reproduction. ### FEC (small) Top-1 Accuracy: 76.74. Result ID: 699883125698665ec4e45402. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"5.44","FLOPs":"1.38","FPS":"854.25±37.49","Memory":"5.2","Type":"Clustering-based"}. This record does not assert independent reproduction. ### CoC (small) Top-1 Accuracy: 76.56. Result ID: 699883125698665ec4e453fd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"14.20","FLOPs":"2.80","FPS":"883.46±29.31","Memory":"3.2","Type":"Clustering-based"}. This record does not assert independent reproduction. ### PVTv2 (b0) Top-1 Accuracy: 75.34. Result ID: 699883125698665ec4e453f4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"3.44","FLOPs":"0.54","FPS":"1195.32±31.63","Memory":"1.7","Type":"Attention-based"}. This record does not assert independent reproduction. ### EfficientViT (m5) Top-1 Accuracy: 75.32. Result ID: 699883125698665ec4e453fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"12.13","FLOPs":"0.52","FPS":"1160.99±44.32","Memory":"1.0","Type":"Attention-based"}. This record does not assert independent reproduction. ### ConvNeXtv2 (N) Top-1 Accuracy: 75.04. Result ID: 699883125698665ec4e453f3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"15.05","FLOPs":"2.46","FPS":"1148.19±28.45","Memory":"2.8","Type":"Convolution-based"}. This record does not assert independent reproduction. ### CoC (tiny) Top-1 Accuracy: 74.9. Result ID: 699883125698665ec4e453fc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"5.28","FLOPs":"1.12","FPS":"887.11±18.92","Memory":"2.2","Type":"Clustering-based"}. This record does not assert independent reproduction. ### ClusterFormer (tiny) Top-1 Accuracy: 73.99. Result ID: 699883125698665ec4e45401. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"30.27","FLOPs":"5.58","FPS":"588.56±5.11","Memory":"9.6","Type":"Clustering-based"}. This record does not assert independent reproduction. ### ClusterFormer Top-1 Accuracy: 73.93. Result ID: 699883125698665ec4e45400. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"15.01","FLOPs":"3.02","FPS":"694.91±5.99","Memory":"12.0","Adjusted":true,"Type":"Clustering-based"}. This record does not assert independent reproduction. ### EfficientViT (m3) Top-1 Accuracy: 73.88. Result ID: 699883125698665ec4e453fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"6.61","FLOPs":"0.26","FPS":"1132.69±24.25","Memory":"1.0","Type":"Attention-based"}. This record does not assert independent reproduction. ### EfficientViT (m2) Top-1 Accuracy: 73.59. Result ID: 699883125698665ec4e453f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"3.99","FLOPs":"0.20","FPS":"1173.13±22.48","Memory":"0.9","Type":"Attention-based"}. This record does not assert independent reproduction. ### ConvNeXtv2 (F) Top-1 Accuracy: 73.14. Result ID: 699883125698665ec4e453f2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"4.89","FLOPs":"0.79","FPS":"1177.66±21.15","Memory":"1.8","Type":"Convolution-based"}. This record does not assert independent reproduction. ### ConvNeXtv2 (A) Top-1 Accuracy: 71.15. Result ID: 699883125698665ec4e453f1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"#Param":"3.42","FLOPs":"0.55","FPS":"1172.04±34.80","Memory":"1.6","Type":"Convolution-based"}. This record does not assert independent reproduction. ### ClusterFormer Top-1 Accuracy: 70.73. Result ID: 699883125698665ec4e453ff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"#Param":"5.63","FLOPs":"1.24","FPS":"669.00±4.52","Memory":"9.3","Adjusted":true,"Type":"Clustering-based"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.