Image Classification on FootpathVision (test)
89AccuracyMobileNetV2
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| MobileNetV2Mode=Transfer, Time / Epoch (s)=0.59, Trainable (MB)=0.01, Frozen (MB)=8.612026.01 | 89 | 89 | 89 | 89 | |
| CustomCNN2026.01 | 86.02 | 86.04 | 85.11 | 85.48 | |
| MobileNetV2Mode=Pretrained, Time / Epoch (s)=2.26, Trainable (MB)=8.49, Frozen (MB)=0.132026.01 | 79 | 85 | 79 | 79 | |
| EfficientNetB0Mode=Pretrained, Time / Epoch (s)=3.29, Trainable (MB)=15.29, Frozen (MB)=0.162026.01 | 75 | 76 | 75 | 75 | |
| Evolved Baseline (Custom CNN)Time / Epoch (s)=126, Trainable (MB)=42.64, Frozen (MB)=0.042026.01 | 71 | 71 | 71 | 71 | |
| EfficientNetB0Mode=Transfer, Time / Epoch (s)=0.81, Trainable (MB)=0.01, Frozen (MB)=15.452026.01 | 63 | 62 | 63 | 62 |