Amer Ouali, Djamel EddineAchour, Abd EselmReporter: Sahraoui, Mohamed2022-10-022022-10-022022-06-10http://dspace.univ-msila.dz:8080//xmlui/handle/123456789/32737Automatic recognition of human emotions has received increasing interest from researchers in the field of computer vision, which has led to the proposal of several methods. Many of them relied on handcrafted features and traditional fusion and classification techniques. The use of deep Learning techniques to automatically extract powerful features from multimedia information as well as their use for merging and classification are new trends that researchers are currently pursuing. In This work, we define a new accurate facial expression detection algorithm based on a deep Learning method, specifically on an intentional convolutional neural network capable of focusing on important parts of the face in an image or video database using more number of needed layers. As a result, our proposed algorithm significantly improves the accuracy rate compared to previously proposed models in database groups FER2013, JAFFE and CK+.enDeep Learning, CNN, Computer Vision, Image recognition.Thesis