Object Localization on Synthetic MNIST 1.0 (train)
0.1413RMSESRL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SRLLoss Function=SRL, Model=CNN → Flatten → Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=20, Trials=302026.06 | 0.1413 | 0.1026 | 0.6118 | |
| SMAELoss Function=SMAE, Model=CNN → Flatten → Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=20, Trials=302026.06 | 0.1427 | 0.1089 | 0.604 | |
| Log CoshLoss Function=Log Cosh, Model=CNN → Flatten → Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=20, Trials=302026.06 | 0.1434 | 0.1099 | 0.5997 | |
| Huber LossLoss Function=Huber Loss, Model=CNN → Flatten → Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=20, Trials=302026.06 | 0.1436 | 0.1115 | 0.5989 | |
| MSELoss Function=MSE, Model=CNN → Flatten → Feedforward Network, Optimizer=Adam, Batch Size=32, Epochs=20, Trials=302026.06 | 0.1447 | 0.1104 | 0.5859 |