Salvato in:
Dettagli Bibliografici
Autori principali: Li, Yixiao, Barth, Julia, Kiefer, Thomas, Fraij, Ahmad
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.07562
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909832273985536
author Li, Yixiao
Barth, Julia
Kiefer, Thomas
Fraij, Ahmad
author_facet Li, Yixiao
Barth, Julia
Kiefer, Thomas
Fraij, Ahmad
contents Multi-modal behavior cloning faces significant challenges due to mode averaging and mode collapse, where traditional models fail to capture diverse input-output mappings. This problem is critical in applications like robotics, where modeling multiple valid actions ensures both performance and safety. We propose EBGAN-MDN, a framework that integrates energy-based models, Mixture Density Networks (MDNs), and adversarial training. By leveraging a modified InfoNCE loss and an energy-enforced MDN loss, EBGAN-MDN effectively addresses these challenges. Experiments on synthetic and robotic benchmarks demonstrate superior performance, establishing EBGAN-MDN as a effective and efficient solution for multi-modal learning tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2510_07562
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle EBGAN-MDN: An Energy-Based Adversarial Framework for Multi-Modal Behavior Cloning
Li, Yixiao
Barth, Julia
Kiefer, Thomas
Fraij, Ahmad
Machine Learning
Multi-modal behavior cloning faces significant challenges due to mode averaging and mode collapse, where traditional models fail to capture diverse input-output mappings. This problem is critical in applications like robotics, where modeling multiple valid actions ensures both performance and safety. We propose EBGAN-MDN, a framework that integrates energy-based models, Mixture Density Networks (MDNs), and adversarial training. By leveraging a modified InfoNCE loss and an energy-enforced MDN loss, EBGAN-MDN effectively addresses these challenges. Experiments on synthetic and robotic benchmarks demonstrate superior performance, establishing EBGAN-MDN as a effective and efficient solution for multi-modal learning tasks.
title EBGAN-MDN: An Energy-Based Adversarial Framework for Multi-Modal Behavior Cloning
topic Machine Learning
url https://arxiv.org/abs/2510.07562