Annonce


[StageM2] Self-Supervised Learning for Epilepsy Detection Under Limited Data

07 Janvier 2026

Catégorie : Postes Stagiaires ;

Telecom SudParis (IP Paris), France

Background

Epilepsy is a chronic neurological disease that poses significant challenges to patients and their families, as such, effective detection and prediction of epilepsy can facilitate patient recovery, streamlining healthcare processes [1]. To perform this detection process, machine learning based methods, with recently deep learning methods allows for accurate detections of epilepsy, commonly by leveraging rich information from bio-signal recorded of the patients, primarily EEG signal serving as primary information [2]. Despite these advances however, there are still challenges persisting on these methods that is the need of the huge amount of data, while it is known that the example of the epilepsy conditions of the patients are quite scarce, with one reason of the different symptoms that each patient may exhibits during the epilepsy seizure [3]. This huge amount of data necessity thus can be a huge drawback, especially if the methods should be tested in the real life conditions [4].

Objective and methodology

This research internship will propose a deep learning methods that can deal with small amount of data. This is done by first evaluating the existing epilepsy seizure detection methods [5,6] under limited data following proper evaluations protocol in machine / deep learning domains[7,8]. Afterwards, the novel methods will be developed primarily by utilizing contrastive [9] learning as pretext task to leverage the unlabeled data, thus increasing methods prediction capacity compared to existing methods. We further also consider to integrate some explainable technique, mainly attention analysis to further shows the methods ability to explain the predictions.

Timeline and Research Lab

  1. The research is expected to be conducted for six months with the results to be communicated in the Master Thesis, internal SAMOVAR event and possibly a publication.
  2. The research will be carried out within the ARTEMIS and EPH department of Télécom SudParis, ARMEDIA team and SAMOVAR laboratory.

Requirements

  1. Knowledge in Deep Learning.
  2. Knowledge in Signal Processing
  3. Knowledge in Machine Learning

Organizations

  • The research will be carried out within the ARTEMIS and EPH department of Télécom SudParis, ARMEDIA team and SAMOVAR laboratory.
  • The working location will be at Evry, France.

———- Applications Requirements ———-

Administration and Technical Requirements

  • Enrolled at Master 2 or in the last year of engineering school.
  • Knowledge in Machine Learning, Deep Learning and Signal Processing.
  • Understanding of the use of Machine Learning (e.g. Scikit-learn), Computer Vision (e.g. OpenCV) and Deep Learning Frameworks, such as (Pytorch or Tensorflow).
  • Prior experience in research is highly advantageous.

Documents to be submitted:

Please send following documents in a single pdf page with the title of ‘Internship-SEL-TSP’ for evaluations to: decky.aspandi_latif@telecom-sudparis.eu or piyush.swami@telecom-suparis.eu :

  • Curriculum Vitae.
  • Current diploma and transcripts.
  • Motivation letter (half to max 1 page, optional).
  • Recommendation letters. (if any)

Application deadlines, selection process and start of the internship:

  • Application deadlines as 31th January 2025 (flexible, first come first served).
  • Selection to be planned from 10th January 2025 onward.
  • Start of work as of 01st February 2025 onward.

Contacts:

Further questions can be addressed to:

References

[1]Swami, Piyush, et al. « A novel robust diagnostic model to detect seizures in electroencephalography. » Expert Systems with Applications 56 (2016): 116-130.

[2]Xu, Jie, et al. « EEG-based epileptic seizure detection using deep learning techniques: A survey. » Neurocomputing 610 (2024): 128644.

[3] Lopes, Fábio, et al. « Addressing data limitations in seizure prediction through transfer learning. » Scientific Reports 14.1 (2024): 14169.

[4] Aspandi, Decky, et al. « Robust facial alignment with internal denoising auto-encoder. » 2019 16th Conference on Computer and Robot Vision (CRV). IEEE, 2019.

[5] Guo, Lianghui, et al. « Clep: Contrastive learning for epileptic seizure prediction using a spatio-temporal-spectral network. » IEEE Transactions on Neural Systems and Rehabilitation Engineering 31 (2023): 3915-3926.

[6] Xiao, Tiantian, et al. « Self-supervised learning with attention mechanism for EEG-based seizure detection. » Biomedical Signal Processing and Control 87 (2024): 105464.

[7] Brigato, Lorenzo, and Luca Iocchi. « A close look at deep learning with small data. » 2020 25th international conference on pattern recognition (ICPR). IEEE, 2021.

[8] Guo, Tianyu, et al. « Contrastive learning from extremely augmented skeleton sequences for self-supervised action recognition. » Proceedings of the AAAI conference on artificial intelligence. Vol. 36. No. 1. 2022.

[9] Antoine Hanna-Asaad, Decky Aspandi-Latif, Titus Zaharia, “MI-Cap: A multi-modal interpretable model for video captioning”, 2025 IEEE International Conference on Content-Based Multimedia Indexing (IEEE CBMI), 22-24th October 2025, Dublin, Ireland

Les commentaires sont clos.