Category: Thèse

Knowledge Graph Embedding & GAN for Medieval Manuscript Studies

We will develop innovative AI techniques to “augment” the historical content of texts and images to study cultural meanings and representations of music and sound in the Middle Ages. Our aim is to develop techniques to both (i) recognize or at least detect textual elements related to performance (words, phrases, etc.) in texts between the 5th and 15th century, and (ii) identify performances (musicians, instruments) in images. The first challenge is to incorporate different data sources lacking explicit contextual linking. Because of the complexity of the subject — the variety of sources, languages,
and artistic contexts — we would like to continue implementing more complete transcultural and diachronic knowledge graphs. Since, knowledge graph embedding techniques have not been exploited in the medieval musicology domain; we will explore different approaches using purely knowledge graph embedding, visual embedding or combined embedding to calculate the similarity between items. Another major challenge in Deep Learning is the need for vast amounts of labeled data for training the images. In
this project, we will use Generative Adversarial Networks (GAN) and Neural Network Style Transfer (NNST) for the generation of large training databases and introduce a pioneering approach towards the generation and completion of historical images containing musical instruments. Furthermore, we will introduce the idea of using GAN for domain transfer between these two domains, i.e., historical representations to “real-life” representations and vice versa. This will be useful for data augmentation as well as image enhancement and completion. Natural Language Processing (NLP) techniques will extract references to musical performances in texts: multilingual for medieval languages and multiple versions of
descriptions are a novel challenge.

Keywords

Knowledge graphs, Ontologies, Cultural Heritage, IIIF, Knowledge graph embedding, Graph Neural Networks, Generative Adversarial Networks (GAN), Neural Network Style Transfer (NNST).

Contact

Dr Xavier Fresquet
Deputy Director of SCAI
phone: + 33 1 44 27 76 05
email: xavier.fresquet@sorbonne-universite.fr
4, Place Jussieu
75005 Paris

PhD position in Computer Science and Numerical Analysis

Deep Neural Networks and Differential Equations

Sorbonne Université, Paris, Fr

Laboratoire Jacques Louis Lions – INRIA Paris and Laboratoire d’informatique de Paris 6

More information :

Available positions

https://mlia.lip6.fr/wp-content/uploads/2020/05/2020-05-Thesis-Proposal-DeepNeuralNetworks-DifferentialEquations.pdf

Advisors and contacts: julien.salomon@inria.fr (Laboratoire Jacques-Louis Lions), patrick.gallinari@lip6.fr (Laboratoire d’informatique de Paris 6)

Starting date : October 2020

Keywords: Machine Learning, Deep Neural Networks, Numerical Analysis, Differential Equations

Differential equations form one of the bedrocks of scientific computing, while neural networks have emerged as the preferred tool of modern machine learning. They offer complementary strengths: the modelling power and interpretability of differential equations, and the approximation and generalization power of deep neural networks.  The objective of the thesis is to develop links between DNNs and DEs in order to start answering central questions like: how could DNNs be used to solve PDEs, how the concepts of numerical analysis could be adapted to DNNs, how to develop hybrid models incorporating both NN modules and ODE/PDE solvers? On the application side, we will focus on PDEs arising from environmental applications. The PhD is at the interplay of machine learning and numerical analysis and will be co-supervised by specialists of the two domains.

Working Environment

The candidate will work at SCAI (Sorbonne Center for Artificial Intelligence) in Paris, under the supervision of Julien Salomon (numerical analysis) and Patrick Gallinari (Machine Learning).

Candidate profile

Master or engineering degree in computer science or applied mathematics. The topic is at the crossroad of machine learning and numerical analysis. The candidate should have a strong scientific background with specialization in one of the two domains, good technical skills in programming. Experience of project development with machine learning platforms such as PyTorch or TensorFlow is a plus.

Application

Send a CV, motivation letter and if possible recommendation letters or contacts to julien.salomon@inria.fr and  patrick.gallinari@lip6.fr