Attached
Beauty cultural activities, such as hairstyles and makeup, require the knowledge of face shape of a
person which, is not always accurate or efficient and requires the expert knowledge. In this paper, we
present a computer aided system based on image processing and machine learning to identify face
shape automatically without having to seek an expert.
The system gets face images as input and then sends them to a pre-trained neural network model. The
network handles the classification tasks and produces the predictions according to the face shape and
gives results for the best matching categories. The network, designed using convolutional neural
network model and a dataset was prepared after pre-processing Google image search results. The
prototype system managed to perform at 98.5 % accuracy level on classifying the shapes of seven
basic face shapes. The experimental results indicate that the proposed approach is a valuable
approach, which can significantly support i
computational effort.
