Optimal and efficient machine learning method for deep semantic segmentation
Four computerized machine learning methods for deep semantic segmentation are fast machine learning method, active machine learning method, optimal machine learning method, and optimal transfer learning method. The fast machine learning method performs a fast deep semantic segmentation learning on training images to generate a deep model. The active machine learning method performs a fast deep semantic segmentation learning on initial training images to generate a first deep model and then an active deep semantic segmentation learning to generate a second deep model. The optimal machine learning method performs a fast deep semantic segmentation learning on initial training images to generate a first deep model and then an optimal deep semantic segmentation learning to generate a second deep model. The optimal transfer learning method applies a pre-trained first deep model on transfer training images and then an optimal deep semantic segmentation transfer learning to generate a second deep model.
1. A computerized optimal machine learning method for deep semantic segmentation, comprising the steps of:
a) inputting a plurality of initial training images and corresponding initial partial ground truth data into electronic storage means;
b) performing a fast deep semantic segmentation learning-1 by electronic computing means using the plurality of initial training images and corresponding initial partial ground truth data to generate a deep model-1;
c) iteratively applying the deep model-1 to the plurality of initial training images to generate initial training image confidence maps; and
d) iteratively performing an optimal deep semantic segmentation learning by electronic computing means using the initial training images, the corresponding initial partial ground truth data and the initial training image confidence maps to generate a deep model-2.
2. The computerized optimal machine learning method of claim 1 , wherein the optimal deep semantic segmentation learning method weights learning loss function by values of the initial training image confidence maps.
3. The computerized optimal machine learning method of claim 1 , wherein the fast deep semantic segmentation learning-1 minimizes a do-not-care robust loss function at an output layer by iteratively adjusting parameters at each layer of the deep model-1.
4. The computerized optimal machine learning method of claim 3 , wherein the parameters are adjusted by a back propagation method.
5. A computerized optimal transfer learning method for deep semantic segmentation, comprising the steps of:
a) inputting a pre-trained semantic segmentation deep model-1 into electronic storage means;
b) inputting at least one transfer training image and corresponding transfer partial ground truth data into electronic storage means;
c) iteratively applying the deep model-1 to the at least one transfer training image to generate transfer training image confidence maps; and
d) iteratively performing an optimal deep semantic segmentation transfer learning by electronic computing means using the at least one transfer training image, the corresponding transfer partial ground truth data and the transfer training image confidence maps to generate a deep model-2.
6. The computerized optimal transfer learning method of claim 5 , wherein the optimal deep semantic segmentation transfer learning method weights learning loss function by values of the transfer training image confidence maps.