Skip to content

GolamMullick/Infant_FacialExpression_Recognition_CNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 

Repository files navigation

Approach: An infant facial expression recognition system of five expressions such as Crying, Smiling, Yawning, Sleeping, Neutral has been developed.The reason for choosing these expressions is because baby mainly shows those basic expressions quite often. For human it’s hard to classify expressions such as surprise, angry, scare and hungry because those expression do not last long. Within a few seconds expressions like surprise can be converted to smiling and angry, scare and hungry to crying. Around five thousands facial images are collected from internet as there is no infant database online and then these images are manually labeled into five classes which has been mentioned earlier. A Pretrained VGGNet Convolutional Neural Network Model is used as Feature Extractor as well as Classifier. Experiments show that Deep CNN is able to classify the images into five expression with accuracy up to 89%.But nevertheless this experiment is still novel and only limited to static image.