search

shivahanifi / SCDD-image-segmentation-keras

Implementation of Segnet, FCN, UNet , PSPNet and other models in Keras.
https://divamgupta.com/image-segmentation/2019/06/06/deep-learning-semantic-segmentation-keras.html
MIT License
0 stars 0 forks source link

determine the exact parameters used for training with scdd #3

Open shivahanifi opened 3 days ago

shivahanifi commented 3 days ago

The models were trained with the same default settings, rather than optimise hyperparameters for each model individually.

Name value
Optimizer adaptive moment (Adam)
learning rate 0.001
Loss function categorical cross-entropy

Parameters I'm not sure about

Name default value used value Notes
steps_per_epoch 512 len(os.listdir(train_image_path)) // batch_size
batch_size 2 2
epochs 5 5 there are notes about around 600 epochs(TBC)