Contact
E-mail: click to reveal
Bio
I completed my master degree in computer science at PolyTech Tours (France) in 2017. After, I worked as a research engineer in the IRISA laboratory in Rennes (France) on the development of geometry learning software during 2 years in IntuiDoc team. Afterwards, I received my PhD in the field of children's handwriting analysis under the supervision of Eric Anquetil (IntuiDoc team) and Elisa Fromont (Lacodam team). I am currently pursuing a Postdoc in Luleå University of Technology (Sweden) on handwriting recognition topic in the Machine learning team.
Keywords: Handwriting Recognition and Segmentation, R-CNN object detector, Seq2Seq, French Children Handwriting, Education
Experiences
Research engineer: geometry learning software
We focus on geometry learning support on tablet for middle-school children. The goal is to propose an e-learning system to help the autonomy of the child with a personalized course. We have designed IntuiGeo, an Intelligent Tutoring System which allows the child to draw freely geometric figures by simulating the traditional pen and paper approach. IntuiGeo interprets, on the fly, the child’s sketches, allowing for a real-time user interaction, and visual, corrective or guidance feedback interventions, in order to enhance the child’s learning performance.
Link for more details on the project
Thesis
Title
Integrating Explicit Knowledge with Deep Learning for Children's Handwriting Recognition and SegmentationJury
Véronique EGLIN - Professeure, INSA de Lyon Harold MOUCHERE - Professeur, Nantes Université Sébastien ADAM - Professeur, INSA de Rouen Nicolas RAGOT - MdC, HDR, Université de Tours Elisa FROMONT - Professeure, Université de Rennes Eric ANQUETIL - Professeur, INSA de RennesAbstract
Our goal is to design a tool for children's handwriting recognition and segmentation in order to accurately analyze the handwriting and provide immediate orthographic feedback to the child. The contributions of this thesis are based on the hybridization of deep learning models with models using explicit expert knowledge. The first contribution consists in integrating the writing dynamics of the online signal in a convolutional neural network for character recognition. The second contribution concerns the improvement of an existing word analysis system. This system uses a guidance mechanism based on the instruction and the words phonetically close to the instruction. It integrates the prediction of a Seq2Seq recognition model into the guidance system. The objective is to overcome the shortcomings of the guidance mechanism when the input words contain non-phonetic errors. The third contribution proposes a new recognition and segmentation framework. It is based on the combination of a model dedicated to recognition and a model dedicated to segmentation. The system also integrates the knowledge contained in the online signal in order to improve the accuracy of the segmentation. Finally, we have developed a rejection mechanism to improve the quality of the feedback given to the child. The results of the experiments demonstrate the interest and efficiency of these contributions.
Publications and communications
PhD
Article in an international peer-reviewed conference
Article in an international peer-reviewed journal
Communication at the SIFED symposium in France without proceeding