arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.02088cs.LGcs.CVstat.ML
收藏

功能Quantization改进了GAN培训

Feature Quantization Improves GAN Training

University at Buffalo, SUNY(纽约州立大学布法罗分校);Microsoft Research(微软研究院)

Yang Zhao, Chunyuan Li, Ping Yu, Jianfeng Gao, Changyou Chen

摘要:

GAN训练的不稳定性是一个长期存在的问题,尽管有显着的研究努力。我们确定,不稳定的问题源于执行小批量统计特征匹配的困难,由于固定的目标分布和逐步生成的分布之间的脆弱平衡。在这项工作中,我们提出了特征量化(FQ)的的离散,嵌入真实和虚假的数据样本到一个共享的离散空间。FQ的量化值被构造为一个进化字典,这是一致的最近的分布历史的特征统计。因此,FQ隐式地实现了紧凑空间中的鲁棒特征匹配。我们的方法可以很容易地插入到现有的GAN模型,在训练中的计算开销很小。我们将FQ应用于9个基准测试的3个代表性GAN模型:用于图像生成的BigGAN,用于人脸合成的StyleGAN GAN,以及用于无监督图像到图像翻译的U-GAT-IT。大量的实验结果表明,所提出的GAN可以在各种任务上大幅提高基线方法的FID分数,实现新的最先进的性能。

英文摘要:

The instability in GAN training has been a long-standing problem despite remarkable research efforts. We identify that instability issues stem from difficulties of performing feature matching with mini-batch statistics, due to a fragile balance between the fixed target distribution and the progressively generated distribution. In this work, we propose Feature Quantization (FQ) for the discriminator, to embed both true and fake data samples into a shared discrete space. The quantized values of FQ are constructed as an evolving dictionary, which is consistent with feature statistics of the recent distribution history. Hence, FQ implicitly enables robust feature matching in a compact space. Our method can be easily plugged into existing GAN models, with little computational overhead in training. We apply FQ to 3 representative GAN models on 9 benchmarks: BigGAN for image generation, StyleGAN for face synthesis, and U-GAT-IT for unsupervised image-to-image translation. Extensive experimental results show that the proposed FQ-GAN can improve the FID scores of baseline methods by a large margin on a variety of tasks, achieving new state-of-the-art performance.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑