Skip to content

Latest commit

 

History

History
55 lines (43 loc) · 1.96 KB

README.md

File metadata and controls

55 lines (43 loc) · 1.96 KB
Raw

🐵 Face Expression Classification using CNN 🚀

FER

• Problem

Identifying the face expression of a human, given an image of him/her.

• Data

Data is taken from Kaggle's Facial Expression Recognition Challenge: https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data

• Evaluation

Evaluation is done based on accuracy and loss between predicted expression and actual expression.

• Features

Some information about the data:

  • We're dealing with images(unstructured data), so its better we use deep learning / transfer learning.
  • Dataset has 3 columns namely emotions, picture, and usage(Training/Testing).
  • Dataset has 35887 rows(images).
  • There are 28709 training images (with column value as Training).
  • There are 3589 testing images (with column value as PublicTest).
  • Dataset has 7 classes (emotions).
    • 0 -> Angry
    • 1 -> Disgust
    • 2 -> Fear
    • 3 -> Happy
    • 4 -> Sad
    • 5 -> Surprise
    • 6 -> Neutral

💚 Developed by Parth Maniar