Image Classification on Food-101N (test)
93.42Top-1 AccuracyLRA-diffusion
Evaluation Results
| Method | Links | |
|---|---|---|
| LRA-diffusionencoder=CLIP2023.05 | 93.42 | |
| SCC (Mixup+GBA)trained from scratch=true, regularization=Mixup, GBA=true2020.08 | 87.55 | |
| SOMNetImageNet pretrained=true2020.08 | 87.5 | |
| SCC (Mixup)trained from scratch=true, regularization=Mixup2020.08 | 87.43 | |
| LongReMixTraining protocol=from pre-trained2021.03 | 87.39 | |
| LongReMix2023.05 | 87.39 | |
| DivideMixTraining protocol=from pre-trained2021.03 | 86.91 | |
| Jigsaw-ViTBackbone=DeiT-Small/16, Number of parameters=22M, FLOPs=4.6G2022.07 | 86.7 | |
| RTEBackbone=ResNet-502021.09 | 86.46 | |
| VRIBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 86.24 | |
| CoDisBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 86.13 | |
| Initial Mixup Modeltrained from scratch=true, regularization=Mixup2020.08 | 86 | |
| WarPIBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 85.91 | |
| WarPIBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 85.9 | |
| ELR+Backbone=ResNet-50, Pre-trained=ImageNet2026.03 | 85.77 | |
| SCC (Vanilla+GBA)trained from scratch=true, regularization=Vanilla, GBA=true2020.08 | 85.76 | |
| PLCBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 85.3 | |
| PLCBackbone=ResNet-50, Pre-training=ImageNet, Trials=32021.03 | 85.28 | |
| PLC2021.06 | 85.28 | |
| PLC2023.05 | 85.28 | |
| PLCBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 85.28 | |
| NRankBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 85.2 | |
| NRankBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 85.2 | |
| Deep Self-LearningImageNet pretrained=true2020.08 | 85.11 | |
| Deep Self LearningBackbone=ResNet-502021.09 | 85.11 | |
| SMPBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 85.11 | |
| DeepSelfTraining protocol=from pre-trained2021.03 | 85.1 | |
| SMPBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 85.1 | |
| DeepSelf2023.05 | 85.1 | |
| MetaCleanerImageNet pretrained=false2020.08 | 85.05 | |
| SCC (Vanilla)trained from scratch=true, regularization=Vanilla2020.08 | 84.87 | |
| MWNetBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 84.72 | |
| MWNetBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 84.7 | |
| Guidance LearningImageNet pretrained=true2020.08 | 84.2 | |
| DeiT-Small/16Backbone=DeiT-Small/16, Number of parameters=22M, FLOPs=4.6G2022.07 | 84.2 | |
| BAREBatch-balance=false2021.06 | 84.12 | |
| BARE2023.05 | 84.12 | |
| Initial Vanilla Modeltrained from scratch=true2020.08 | 84.08 | |
| CleanNetImageNet pretrained=true2020.08 | 83.95 | |
| CleanNetBackbone=ResNet-50, Pre-training=ImageNet2021.03 | 83.95 | |
| CleanNet (soft)Backbone=ResNet-502021.09 | 83.95 | |
| CleanNet2021.05 | 83.95 | |
| CleanNetTraining protocol=from pre-trained2021.03 | 83.95 | |
| CleanNet2021.06 | 83.95 | |
| CleanNet2023.05 | 83.95 | |
| CleanNetBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 83.5 | |
| CleanNet (hard)Backbone=ResNet-502021.09 | 83.47 | |
| CNethBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 83.47 | |
| CEBackbone=ResNet-50, Number of parameters=25.6M, FLOPs=4G2022.07 | 81.7 | |
| StandardBackbone=ResNet-50, Pre-training=ImageNet2021.03 | 81.67 | |
| CE2021.05 | 81.67 | |
| CCEvariant=standard2021.06 | 81.67 | |
| Standard2023.05 | 81.67 | |
| Base ModelBackbone=ResNet-50, Pre-trained=ImageNet2026.03 | 81.67 | |
| BaselineBackbone=ResNet-502021.09 | 81.44 | |
| Cross-EntropyTraining protocol=from pre-trained2021.03 | 81.44 | |
| LongReMixTraining protocol=from scratch2021.03 | 78.57 | |
| RGC2026.05 | 78.5 | |
| DivideMixTraining protocol=from scratch2021.03 | 75.53 | |
| IDORef.=NeurIPS252026.05 | 74.4 | |
| EarlyCutRef.=NeurIPS252026.05 | 72.1 | |
| DCDRef.=ICCV252026.05 | 71.2 | |
| PSSCLRef.=PR252026.05 | 70.7 | |
| UNICONRef.=CVPR222026.05 | 70.4 | |
| ProMixRef.=IJCAI232026.05 | 70.1 | |
| RRLRef.=ICCV212026.05 | 69.7 | |
| L2BRef.=CVPR242026.05 | 68.9 | |
| LongReMixRef.=PR232026.05 | 68.7 | |
| DivideMixRef.=ICLR202026.05 | 67.2 |