Mohamed Dhafer Hadj Amor

Bidirectional recurrent model implementation for Arabic word recognition that aims to find the word features as personal pronouns, root, scheme, tense and is the word a verb or a noun based on images. The importance of this project came out from the rarity of the subject and especially from the lack of Arabic data that's why we tried to make our proper data, then I worked on very interesting models such as CRNN and BiLSTM. Keywords: sequence modeling, RNN, LSTM, BiLSTM, deep learning, Named Entity, NLP.

-quality study with the Six Sigma method (financial study, technical, legal). -Brainstorming in collaboration with the quality team and the maintenance team. -Teamwork by dividing tasks. During this internship I learned how to integrate myself into such an environment, Therefore I learned how to work in a team, make constructive feedback and how to be diplomatic in order to resolve conflicts and disagreements.