Image Classification on Aircrafts
87.27Top-1 AccuracyBYOL+CVSA
Evaluation Results
| Method | Links | |
|---|---|---|
| BYOL+CVSAStage 2 pre-training=true, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 87.27 | |
| SupervisedStage 2 pre-training=false, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 87.25 | |
| BYOLStage 2 pre-training=false, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 87.19 | |
| BYOL+DiLoStage 2 pre-training=true, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 87.03 | |
| InsLocStage 2 pre-training=true, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 86.6 | |
| SimCLRStage 2 pre-training=false, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 85.96 | |
| MoCo.v2Stage 2 pre-training=false, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 85.49 | |
| BYOLStage 2 pre-training=true, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 84.58 | |
| MoCo.v2Stage 2 pre-training=true, Evaluation protocol=fine-tuned evaluation, Stage 1 pre-training dataset=ImageNet-1k, Backbone=ResNet-50, Pre-training epochs=8002021.06 | 83.25 | |
| DeiT-TGFLOPS=1.262023.05 | 78.7 | |
| Zero-TPGFLOPS=0.91, Backbone=DeiT-T2023.05 | 76.7 | |
| ATSGFLOPS=0.90, Backbone=DeiT-T2023.05 | 76 | |
| ToMeGFLOPS=0.90, Backbone=DeiT-T2023.05 | 75.2 | |
| LOUPEmode=zero-shot2022.08 | 54.9 | |
| CLIPmode=zero-shot2022.08 | 36.1 | |
| OpenAI-CLIP ViT-LN=400M, Zero-shot=true2024.09 | 36.1 | |
| OpenAI-CLIP VIT-LN=400M, Zero-shot=true2024.09 | 36.1 | |
| DINOv2-ARLN=20M, Zero-shot=true2024.09 | 28.1 | |
| DINOv2-ARLN=20M, Zero-shot=true2024.09 | 28.1 | |
| LAION-CLIP ViT-LN=400M, Zero-shot=true2024.09 | 24.6 | |
| LAION-CLIP VIT-LN=400M, Zero-shot=true2024.09 | 24.6 | |
| DINOv2-MpNetN=20M, Zero-shot=true2024.09 | 23 | |
| DINOv2-MpNetN=20M, Zero-shot=true2024.09 | 23 | |
| LiT L16LN=112M, Zero-shot=true2024.09 | 15.2 | |
| LiT L16LN=112M, Zero-shot=true2024.09 | 15.2 | |
| LiTDA-baseN=0.5M, Zero-shot=true2024.09 | 1.7 | |
| LiTLwA-baseN=0.5M, Zero-shot=true2024.09 | 1.6 |