{"ID":5675247,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-07T01:06:03.009715918Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01907","arxiv_id":"2607.01907","title":"Population-Based Multi-Objective Training of Discriminators for Semi-Supervised GANs","abstract":"Semi-supervised generative adversarial networks (SSL-GANs) can exploit large unlabeled datasets while retaining a classifier in the discriminator, but their training is often unstable. This paper proposes a population-based evolutionary training strategy in which discriminator learning is formulated as a multi-objective optimization problem. Instead of aggregating the supervised and unsupervised components of the SSL objective into a single scalar loss, the method maintains a population of discriminators ranked by Pareto dominance, enabling the exploration of different trade-offs between classification accuracy and real/fake discrimination. This formulation aims to improve both roles of SSL-GANs: learning accurate classifiers and training generators capable of producing realistic samples. We analyze several variants, including an elitist strategy and a mono-objective ablation, to assess the role of multi-objective selection. Experiments on MNIST with limited labels show improved training robustness compared to SSL-GAN and CE-SSL-GAN state-of-the-art baselines, while the elitist variant consistently achieves the highest classification accuracy.","short_abstract":"Semi-supervised generative adversarial networks (SSL-GANs) can exploit large unlabeled datasets while retaining a classifier in the discriminator, but their training is often unstable. This paper proposes a population-based evolutionary training strategy in which discriminator learning is formulated as a multi-objectiv...","url_abs":"https://arxiv.org/abs/2607.01907","url_pdf":"https://arxiv.org/pdf/2607.01907v1","authors":"[\"Francisco Sedeño\",\"Francisco Chicano\",\"Jamal Toutouh\"]","published":"2026-07-02T09:06:06Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.CV\"]","methods":"[\"LoRA\",\"Generative Adversarial Network\"]","has_code":false}