paper-with-me

홈 › Papers

Hellinger loss function for Generative Adversarial Networks

2025-12-13 · Giovanni Saraceno, Anand N. Vidyashankar, Claudio Agostinelli arxiv

We propose Hellinger-type loss functions for training Generative Adversarial Networks (GANs), motivated by the boundedness, symmetry, and robustness properties of the Hellinger distance. We define an adversarial objective based on this divergence and study its statistical properties within a general parametric framework. We establish the existence, uniqueness, consistency, and joint asymptotic normality of the estimators obtained from the adversarial training procedure. In particular, we analyze the joint estimation of both generator and discriminator parameters, offering a comprehensive asymptotic characterization of the resulting estimators. We introduce two implementations of the Hellinger-type loss and we evaluate their empirical behavior in comparison with the classic (Maximum Likelihood-type) GAN loss. Through a controlled simulation study, we demonstrate that both proposed losses yield improved estimation accuracy and robustness under increasing levels of data contamination.

📄 PDF Abstract BibTeX arXiv:2512.12267

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

ANROT-HELANet: Adverserially and Naturally Robust Attention-Based Aggregation Network via The Hellinger Distance for Few-Shot Classification

2025-09-14 · Gao Yu Lee, Tanmoy Dam, Md Meftahul Ferdaus, Daniel Puiu Poenar 외 arxiv

Few-Shot Learning (FSL), which involves learning to generalize using only a few data samples, has demonstrated promising and superior performances to ordinary CNN methods. While Bayesian based estimation approaches using…

Image ReconstructionFew-Shot Learning

Cumulant GAN

2020-06-11 · Yannis Pantazis, Dipjyoti Paul, Michail Fasoulakis, Yannis Stylianou 외

In this paper, we propose a novel loss function for training Generative Adversarial Networks (GANs) aiming towards deeper theoretical understanding as well as improved stability and performance for the underlying optimiz…

Image Generation

Fast Rates for General Unbounded Loss Functions: from ERM to Generalized Bayes

2016-05-01 · Peter D. Grünwald, Nishant A. Mehta

We present new excess risk bounds for general unbounded loss functions including log loss and squared loss, where the distribution of the losses may be heavy-tailed. The bounds hold for general estimators, but they are o…

Bayesian Inference

Private Minimum Hellinger Distance Estimation via Hellinger Distance Differential Privacy

2025-01-24 · Fengnan Deng, Anand N. Vidyashankar

Objective functions based on Hellinger distance yield robust and efficient estimators of model parameters. Motivated by privacy and regulatory requirements encountered in contemporary applications, we derive in this pape…

Hellinger-Kantorovich Gradient Flows: Global Exponential Decay of Entropy Functionals

2025-01-28 · Alexander Mielke, Jia-Jie Zhu

We investigate a family of gradient flows of positive and probability measures, focusing on the Hellinger-Kantorovich (HK) geometry, which unifies transport mechanism of Otto-Wasserstein, and the birth-death mechanism of…

Inference Optimization