Julie Mordacq

T-REGS: Minimum Spanning Tree Regularization for Self-Supervised Learning

Julie Mordacq, David Loiseaux, Vicky Kalogeiton, Steve Oudot

Self-supervised learning (SSL) has emerged as a powerful paradigm for learning representations without labeled data, often by enforcing invariance to input transformations such as rotations or blurring. Recent studies have highlighted two pivotal properties for effective representations: (i) avoiding dimensional collapse-where the learned features occupy only a low-dimensional subspace, and (ii) enhancing uniformity of the induced distribution. In this work, we introduce T-REGS, a simple regularization framework for SSL based on the length of the Minimum Spanning Tree over the learned representation.

Spotlight
Proc. Conference on Neural Information Processing Systems (NeurIPS), 2025

ADAPT: Multimodal Learning for Detecting Physiological Changes under Missing Modalities

Julie Mordacq, Leo Milecki, Maria Vakalopoulou, Steve Oudot, Vicky Kalogeiton

Multimodality has recently gained attention in the medical domain, where imaging or video modalities may be integrated with biomedical signals or health records. Yet, two challenges remain: balancing the contributions of modalities, especially in cases with a limited amount of data available, and tackling missing modalities. To address both issues, in this paper, we introduce the AnchoreD multimodAl Physiological Transformer (ADAPT), a multimodal, scalable framework with two key components: (i) aligning all modalities in the space of the strongest, richest modality (called anchor) to learn a joint embedding space, and (ii) a Masked Multimodal Transformer, leveraging both inter- and intra-modality correlations while handling missing modalities.

Multimodal Learning for Detecting Stress under Missing Modalities

Julie Mordacq, Leo Milecki, Maria Vakalopoulou, Steve Oudot, Vicky Kalogeiton

Dealing with missing modalities is critical for many real-life applications. In this work, we propose a scalable framework for detecting stress induced by specific triggers in multimodal data with missing modalities. Our method has two key components: (i) aligning all modalities in the space of the strongest modality (the video) for learning a joint embedding space and (ii) a Masked Multimodal Transformer, leveraging inter- and intra-modality correlations while handling missing modalities. We validate our method through experiments on the StressID dataset, where we set the new state of the art while demonstrating its robustness across various modality scenarios and its high potential for real-life applications.