Research Article
Unsupervised Domain Adaptation for Facial Expression Recognition Using Generative Adversarial Networks
Algorithm 1
Fine-tuning process using DPL.
Input: | : pretrained CNN model | : source images | : GAN generated images | : training images | Functions: | : probability output of given | : calculate the loss using DPL loss function | : calculate the loss using cross-entropy loss function | Training: | (1) for each epoch: | (2) for each : | (3) | (4) if : | (5) | (6) else if : | (7) | (8) Update with | (9) end | (10) end |
|