Hi,
Thank you very much for providing the code.
While I am training the human gaze models, I noticed that you did not use GH_normalization() function to normalize the true heatmap, so that the sum of all values in each heatmap is 1. Which matchs the outputs of softmax function of the human gaze network.
Can you give me some details, how you process the human gaze dataset into heatmap to train the model?
Thank you very much for your time.
Hi,
Thank you very much for providing the code.
While I am training the human gaze models, I noticed that you did not use GH_normalization() function to normalize the true heatmap, so that the sum of all values in each heatmap is 1. Which matchs the outputs of softmax function of the human gaze network.
Can you give me some details, how you process the human gaze dataset into heatmap to train the model?
Thank you very much for your time.