Research milestones, publications, awards, and group updates. Publication announcements link to the corresponding bibliographic record whenever one is available.
We are recruiting a PhD student for the Hi! PARIS project SIGMA, “Signal-aware Graph Summarization with GNN Guarantees”, co-supervised with Aref Einizade and Philippe Ciblat at Télécom Paris and Télécom SudParis. The three-year position starts in spring 2027; applications are reviewed on a rolling basis.
From September to October 2026, I was a visiting professor at Universidad de Antioquia, Colombia, where I taught an intensive course on Machine Learning with Graphs.
Our paper “A Sparse Low-Rank Biclique Decomposition for Graphs” was accepted at NeurIPS 2026.
Our paper “CALICE: Continuous bitrate control with Adapted LIC modEl” was accepted in ACM Transactions on Multimedia Computing, Communications, and Applications.
I was selected as an ICML 2026 Gold Reviewer, recognizing the top 25% of reviewers.
The papers study feature-aware hypergraph generation and graph-informed flow matching for spatiotemporal imputation.
I was selected among the Top 200 Reviewers at ICLR 2026.
Our work on stability-refined graph learning for soundscape connectomes was accepted at IJCNN 2026.
Our paper on generalization bounds for spectral graph neural networks was accepted at AISTATS 2026.
Our wildlife image representation model for geographical domain shift was accepted in the International Journal of Computer Vision.
Our work on transferability estimation for pruned backbones was accepted in Neurocomputing.
One main-track and three workshop papers cover simplicial neural networks, tensorial PDEs, materials science, and environmental applications.
I became a member of the European Laboratory for Learning and Intelligent Systems Society.
Our work on continuous neural models for simplicial complexes was accepted at NeurIPS 2025.
Our semi-supervised graph learning method for underwater source localization was accepted in the Journal of the Acoustical Society of America.
Our paper on Subgraph Gaussian Embedding Contrast was accepted at ECML PKDD 2025.
Our Piecewise Constant Spectral Graph Neural Network paper was accepted in Transactions on Machine Learning Research.
Our diffusion model for point-cloud compression at low bit-rates was accepted at ICME 2025.
Our fused Gromov–Wasserstein approach to subgraph contrastive learning was accepted in TMLR.
WiGNet and CATALOG were selected for oral presentation at WACV 2025.
Our graph-based semi-supervised method was accepted in Computer Vision and Image Understanding.
We received an NVIDIA Academic Grant for research connecting foundation models and camera-trap recognition.
Our diffusion-based hypergraph generation method was accepted at AAAI 2025.
Our camera-trap language-guided contrastive learning model was accepted at WACV 2025.
Our paper on efficient higher-order graph neural networks was accepted in IEEE Transactions on Signal and Information Processing over Networks.
Our work on continuous product graph neural networks was accepted at NeurIPS 2024.