Image Classification on CIFAR10 (Accuracy and Efficiency)
94.37AccuracyBP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BPBackbone=ResNet34, # Layers=42025.09 | 94.37 | 49 | 0.12 | |
| BPBackbone=ResNet18, # Layers=42025.09 | 94.01 | 61.25 | 0.09 | |
| HOSVDBackbone=ResNet34, # Layers=42025.09 | 93.8 | 0.92 | 12.25 | |
| LANCEBackbone=ResNet34, # Layers=42025.09 | 93.56 | 1.01 | 0.06 | |
| BPBackbone=ResNet34, # Layers=22025.09 | 93.48 | 24.5 | 0.06 | |
| HOSVDBackbone=ResNet18, # Layers=42025.09 | 93.16 | 1.16 | 15.56 | |
| LANCEBackbone=ResNet18, # Layers=42025.09 | 92.89 | 1.11 | 0.05 | |
| HOSVDBackbone=ResNet34, # Layers=22025.09 | 92.7 | 0.42 | 6.12 | |
| BPBackbone=ResNet18, # Layers=22025.09 | 92.69 | 24.5 | 0.05 | |
| LANCEBackbone=ResNet34, # Layers=22025.09 | 92.62 | 0.46 | 0.03 | |
| BPBackbone=MobileNetV2, # Layers=42025.09 | 92.23 | 57.42 | 0.01 | |
| HOSVDBackbone=ResNet18, # Layers=22025.09 | 92.17 | 0.71 | 6.12 | |
| ASIBackbone=ResNet18, # Layers=42025.09 | 91.9 | 1.89 | 0.06 | |
| LANCEBackbone=ResNet18, # Layers=22025.09 | 91.8 | 0.75 | 0.03 | |
| BPBackbone=MobileNetV2, # Layers=22025.09 | 91.72 | 30.62 | 0.01 | |
| ASIBackbone=ResNet34, # Layers=42025.09 | 91.54 | 1.24 | 0.07 | |
| ASIBackbone=ResNet18, # Layers=22025.09 | 90.77 | 1.24 | 0.04 | |
| ASIBackbone=ResNet34, # Layers=22025.09 | 90.09 | 0.49 | 0.03 | |
| HOSVDBackbone=MobileNetV2, # Layers=42025.09 | 89.03 | 0.16 | 22.86 | |
| LANCEBackbone=MobileNetV2, # Layers=42025.09 | 88.99 | 0.17 | 0.01 | |
| ASIBackbone=MobileNetV2, # Layers=42025.09 | 88.03 | 0.63 | 0.06 | |
| HOSVDBackbone=MobileNetV2, # Layers=22025.09 | 85.67 | 0.13 | 11.87 | |
| ASIBackbone=MobileNetV2, # Layers=22025.09 | 85.09 | 0.26 | 0.01 | |
| LANCEBackbone=MobileNetV2, # Layers=22025.09 | 84.16 | 0.15 | 0.01 | |
| BPBackbone=MCUNet, # Layers=42025.09 | 82.91 | 17.57 | 0.001 | |
| LANCEBackbone=MCUNet, # Layers=42025.09 | 76.23 | 0.09 | 0 | |
| BPBackbone=MCUNet, # Layers=22025.09 | 75.85 | 11.71 | 0 | |
| HOSVDBackbone=MCUNet, # Layers=42025.09 | 75.63 | 0.12 | 2.59 | |
| HOSVDBackbone=MCUNet, # Layers=22025.09 | 70.91 | 0.04 | 1.93 | |
| LANCEBackbone=MCUNet, # Layers=22025.09 | 70.64 | 0.04 | 0 |