Deep learning and tensor decomposition for the analysis of patterns in signals and multimodal imaging. Application to neuropathies
Deep learning and tensor decomposition for the analysis of patterns in signals and multimodal imaging. Application to neuropathies
- context
- IBISC, équipe SIAM, LISSI
- Contexte et objectifs
Apply for this position
Informations générales
Le Laboratoire IBISC (Informatique, Bioinformatique, Systèmes Complexes EA 4526) est un laboratoire de l’Université d’Évry Paris-Saclay structuré en quatre équipes de recherche : AROBAS, COSMO, IRA2 et SIAM. Une particularité du laboratoire est sa recherche pluridisciplinaire ainsi que sa localisation sur deux sites de l’université : IBGBI et PELVOUX. Cette spécificité est également renforcée par son rattachement à deux UFRs scientifiques distinctes : l’UFR Sciences Fondamentales et Applications (SFA) et l’UFR Science et Technologie (ST). Le laboratoire IBISC développe résolument une stratégie de collaboration et de valorisation de la recherche avec l’industrie ainsi qu’une stratégie de recherche ouverte à l’international. En 2023, le laboratoire IBISC a accueilli 23% du personnel enseignant et de recherche de l’UEVE qui porte plusieurs responsabilités aussi bien à l’université d’Évry (LMD, UFRs, IUT, VPs) qu’à l’université de Paris-Saclay (Graduate schools en Informatique et Sciences du Numérique (ISN) et en Sciences de l’Ingénierie et des Systèmes (SIS)).
Le Laboratoire LISSI (Laboratoire Images, Signaux et Systèmes Intelligents) est une équipe d’accueil de l’Université Paris Est Créteil (EA N° 3956). Il a été créé en janvier 2005 suite à la fusion de trois équipes de l’UPEC : LIIA, LERISS, et I2S. Le laboratoire est rattaché à l’Institut Universitaire de Technologie de Créteil-Vitry et accueille des enseignants-chercheurs de 4 Unités de Formation et de Recherche de l’UPEC. Le LISSI développe des recherches multidisciplinaires, théoriques et appliquées, dans le domaine des sciences et technologies de l’information et de la communication et en particulier de l’intelligence artificielle. Les applications traitées se situent principalement dans le domaine des technologies pour la santé et le bien-être et sont axées sur des problématiques difficiles telles l’aide au diagnostic et au suivi thérapeutique, le vieillissement, l’assistance aux personnes dépendantes ou handicapées et la e-santé. Le projet scientifique du laboratoire s’articule autour des thèmes de recherche suivants :
- L’optimisation difficile, et l’imagerie vasculaire;
- La perception bio-inspirée de l’environnement, et la biométrie cachée;
- La robotique d’assistance et l’intelligence ambiante;
- Le contrôle adaptatif des systèmes de communication.
- Basic AI and Data Science : classification from image features
- Specialized ML and AI : signal, image, vision
- Application domain : AI-based imagery by MFI
- Mots-clés : deep learning, tensor decomposition, machine learning, deep tech, source separation, tensor decomposition, time series, neuroimaging, stroke
Context et objectives
The exponential development of AI and neural networks is renewing the study of time series from both a fundamental and applied point of view. Particularly for multivariate signals, the tensor can be a more adequate representation than the matrix, because it avoids the loss of data structure, and therefore the loss of information. Automatic learning on tensor data is classically carried out by linear tensor decomposition, for example CPD/PARAFAC or Tucker [Sid17]. Recently, tensor representations have been integrated into neural networks and have enabled significant developments in deep learning, particularly in the field of images, by reducing the number of parameters to be estimated.
To increase the identifiability and interpretability of deep neural models, constraints are added, for example non-negativity, classic in a matrix and tensor learning framework [Kol08]. In deep learning, variational autoencoders have been interpreted in a non-negative matrix factorization framework, but also as a CPD tensor factorization, and even non-negative Tucker [Mar22]. Autoencoders belong to the family of generative models. They make it possible to discover latent spaces by learning an automorphism x=f(x). Their latent space can be structured in tensor form, which provides very good performance [Pan21]. It has been shown that this allows a compromise in terms of performance and interpretability, between a simple unconstrained autoencoder and a non-negative Tucker model, for different tasks (segmentation, pattern detection). However, this preliminary work leaves significant room for progress, and the properties of this type of hybrid model are still poorly understood.
Work program
First of all, we will establish a benchmark of the different approaches. Then we will modify the constraints which structure the tensor decomposition in an auto-encoder/Tucker decomposition type model. We will evaluate and compare the characteristics of several architectures for the autoencoder. The proposed algorithms will be tested on data from several application fields currently examined in our respective laboratories: powers transmitted on an electricity transmission network; calibration of pollutant sensors; prediction of sports performance, segmentation of brain tumors. This work could continue in thesis (1) by comparing the performances of the representation in the temporal, time-frequency, time-scale domains (2) by applying these tensor decompositions on Boltzmann machines (DB networks and diffusion model) (3) by studying the influence of the network structure of the underlying phenomenon on the signal representation. Industrial collaborations are possible.
References
[Kol08] Kolda, Bader, « Tensor decompositions and applications », in: SIAM review 51.3 (2009), pp. 455–500.
[Sid17] Sidiropoulos et al. « Tensor Decomposition for Signal Processing and Machine Learning »IEEE Transactions on Signal Processing, 2017.
[Pan21] Panagakis et al. « Tensor Methods in Computer Vision and Deep Learning » Proceedings of the IEEE, https://doi.org/10.1109/JPROC.2021.3074329
[Mar22] Marmoret, « Unsupervised Machine Learning Paradigms for the Representation of Music Similarity and Structure », thèse IMT Atlantique, 2022.
Ability to understand and develop adaptive learning algorithms and process data, index them and exploit them in an operational system to achieve the mission described above.
Programming skills: Python or C/C++. Experience with Tensorflow or Pytorch would be a plus.
French is not mandatory. Professional qualities sought after autonomy, personal skills to interact with research and business teams, motivation for new technologies, creativity to implement an innovative solution. Scientific and material supervision and conditions The project is multidisciplinary, at the interface of machine learning, computer science and physics. The student will be supervised by Vincent Vigneron and Aurelien Hazan, both specialists in machine learning, signal and image processing. The Synapse team (LISSI) develops tools for analyzing network phenomena. The IBISC SIAM team is recognized for its expertise in deep learning.
Remuneration: €3600/6 months
Contact: send resume and grades of bachelor+master to Aurélien Hazan and Vincent Vigneron.
Phone: +33 6 63 568 760