FacialExpresseonDetection

---

• Project Title : Facial Expression Recognition using Convolutional Neural Network *

---

•   		Author : A. Peter Hudson	      		   *
  

---

---

============================================================

The Project folder Contains the following files and folders:

============================================================

1. CNN_model.py

2. DetectFacialExpression.py

3. FaceRecognitionUtility.py

4. TestImages.py

5. TrainDataset.py

6. haar_cascade_frontalface_default.xml

7. model.h5

8. Dataset - zip file

9. TestData - zip file

10. Weights - Folder

11. DatasetGenerator - Folder

============================================================

=========================

• File/Folder Description *

=========================

================================================================================================================================

1. CNN_model.py : A python file containing the code of the architecture of the CNN. The file is used as a module.

2. DetectFacialExpression.py : Python program for the real-time facial expression detection.

   Command to run/execute the program - "python DetectFacialExpression.py"

3. FaceRecognitionUtility.py : Python program contains functions used for pre-processing of image. The file is used as a module.

4. TestImages.py: Python program to test the CNN model on images.

   Command to run/execute the program - "python TestImages.py"

   Note: Make sure the Directory path is correct in the program.

5. TrainDataset.py: Python program to train datasets to update weights for the model.

   Command to run/execute the program - "python TrainDataset.py"

   Note: Make sure the Directory path is correct in the program.

6. haar_cascade_frontalface_default.xml: The xml file used in face detection in an image. The file is used as a module.

7. model.h5: This file contains the precalculated or the newly created weights which is the output of the "TrainDataset.py".

8. Dataset.zip :
   This zip file contains six directories of 6 different emotions namely - Anger, Happy, Sad, Surprise, Fear, Neutral.

   Each directory contains images under that category.

   Used for training the CNN.

9. TestData.zip : This directory contains six directories of 6 different emotions namely - Anger, Happy, Sad, Surprise, Fear, Neutral.

   Each directory contains images under that category.

   Used for testing the CNN.

10. Weights - Folder

    Contains weights obtained as the size of the dataset was increased.

11. DatasetGenerator - Folder

    Contains two files.

    1. fer2013.csv : This file contains pixel values of the images used for training.

    2. reconstruct.py : Python program to gnerate the images from fer2013.csv file.

===================================================================================================================================