Portrait de Luca Scimeca

Luca Scimeca

Postdoctorat - UdeM
Superviseur⋅e principal⋅e
Yoshua Bengio
Sujets de recherche
Apprentissage de représentations
Apprentissage profond
Biologie computationnelle
Causalité
Inférence bayésienne
Modèles génératifs
Modèles probabilistes
Site web
Google Scholar
GitHub

Publications

Leveraging Diffusion Disentangled Representations to Mitigate Shortcuts in Underspecified Visual Tasks
Luca Scimeca
Alexander Rubinstein
Armand Mihai Nicolicioiu
Damien Teney
Yoshua Bengio
Spurious correlations in the data, where multiple cues are predictive of the target labels, often lead to shortcut learning phenomena, where… (voir plus) a model may rely on erroneous, easy-to-learn, cues while ignoring reliable ones. In this work, we propose an ensemble diversification framework exploiting the generation of synthetic counterfactuals using Diffusion Probabilistic Models (DPMs). We discover that DPMs have the inherent capability to represent multiple visual cues independently, even when they are largely correlated in the training data. We leverage this characteristic to encourage model diversity and empirically show the efficacy of the approach with respect to several diversification objectives. We show that diffusion-guided diversification can lead models to avert attention from shortcut cues, achieving ensemble diversity performance comparable to previous methods requiring additional data collection.
2023-10-03
ArXiv (prépublication)
doi.org
arxiv.org