denseNet121Prediction
fun denseNet121Prediction()
Content copied to clipboard
This examples demonstrates the inference concept on DenseNet121 model:
Model configuration, model weights and labels are obtained from TFModelHub.
Weights are loaded from .h5 file, configuration is loaded from .json file.
Model predicts on a few images located in resources.
Special preprocessing (used in DenseNet121 during training on ImageNet dataset) is applied to the images before prediction.
NOTE: Input resolution is 224*224