用证据理论提升GAN多样性，让生成结果更丰富可靠。
Epistemic Generative Adversarial Networks
- 基于证据理论改进GAN损失函数，同时优化生成器和判别器。
- 生成器可预测每个像素的置信度分布，量化输出不确定性。
- 在多样性和代表性上显著提升，适合需要可信生成的场景。
生成模型，尤其是生成对抗网络（GAN），常因输出多样性不足而生成重复样本。本文提出一种基于达摩-谢弗证据理论的GAN损失函数新范式，应用于生成器与判别器。此外，我们对生成器结构进行改进，使其能为每个图像像素预测一个质量函数，从而量化输出不确定性，并利用该不确定性生成更具多样性和代表性的样本。实验表明，该方法不仅提升了生成多样性，还为生成过程中的不确定性建模与解释提供了严谨框架。
原文摘要 · Abstract (English)
Generative models, particularly Generative Adversarial Networks (GANs), often suffer from a lack of output diversity, frequently generating similar samples rather than a wide range of variations. This paper introduces a novel generalization of the GAN loss function based on Dempster-Shafer theory of evidence, applied to both the generator and discriminator. Additionally, we propose an architectural enhancement to the generator that enables it to predict a mass function for each image pixel. This modification allows the model to quantify uncertainty in its outputs and leverage this uncertainty to produce more diverse and representative generations. Experimental evidence shows that our approach not only improves generation variability but also provides a principled framework for modeling and interpreting uncertainty in generative processes.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。