From: Identification of difficult laryngoscopy using an optimized hybrid architecture
Models | Location | Test accuracy | Average precision score |
---|---|---|---|
CNN+SPP_1 | Location 1: After three convolutional layers | 82.93% | 82.93% |
CNN+SPP_2 | Location 2: After each convolutional layer, before Batch Normalization | 83.33% | 83.74% |
CNN+SPP_3 | Location 3: After each convolutional layer, after Batch Normalization | 83.26% | 83.20% |