7 papers
Learning Common and Salient Generative Factors Between Two Image Datasets
Yunlong He, Gwilherm Lesné, Ziqian Liu +2
Recent advancements in image synthesis have enabled high-quality image generation and manipulation. Most works focus on: 1) conditional manipulation, where an image is modified con…
SAGA: Learning Signal-Aligned Distributions for Improved Text-to-Image Generation
Paul Grimal, Michaël Soumm, Hervé Le Borgne +2
State-of-the-art text-to-image models produce visually impressive results but often struggle with precise alignment to text prompts, leading to missing critical elements or uninten…
Fairer Analysis and Demographically Balanced Face Generation for Fairer Face Verification
Alexandre Fournier-Montgieux, Michael Soumm, Adrian Popescu +2
Face recognition and verification are two computer vision tasks whose performances have advanced with the introduction of deep representations. However, ethical, legal, and technic…
Causal Inference Tools for a Better Evaluation of Machine Learning
Michaël Soumm
We present a comprehensive framework for applying rigorous statistical techniques from econometrics to analyze and improve machine learning systems. We introduce key statistical me…
Quantifying User Coherence: A Unified Framework for Analyzing Recommender Systems Across Domains
Michaël Soumm, Alexandre Fournier-Montgieux, Adrian Popescu +1
The performance of Recommender Systems (RS) varies significantly across users, yet the underlying reasons for this variance remain poorly understood. This paper introduces a unifie…
Toward Fairer Face Recognition Datasets
Alexandre Fournier-Montgieux, Michael Soumm, Adrian Popescu +2
Face recognition and verification are two computer vision tasks whose performance has progressed with the introduction of deep representations. However, ethical, legal, and technic…
An Analysis of Initial Training Strategies for Exemplar-Free Class-Incremental Learning
Grégoire Petit, Michael Soumm, Eva Feillet +4
Class-Incremental Learning (CIL) aims to build classification models from data streams. At each step of the CIL process, new classes must be integrated into the model. Due to catas…