论文标题
缩小去量化差距：pixelcnn作为单层流
Closing the Dequantization Gap: PixelCNN as a Single-Layer Flow
论文作者
论文摘要
流模型最近在建模序数离散数据（例如图像和音频）方面取得了长足进展。由于流量模型的连续性，将其用于此类离散数据时通常会进行取消化，从而导致了可能性的下限估计。在本文中，我们介绍了子集流，这是一类可以易于改变有限体积的流，从而可以精确地计算出离散数据的可能性。基于子集流，我们将序数离散自回旋模型（包括波烯，Pixelcnns和Transformers）确定为单层流。我们使用流程公式比较经过训练和评估的模型与确切的可能性或其去量化下限。最后，我们研究了由像素和非运动耦合层组成的多层流，并在CIFAR-10上证明了对经过去量化训练的流程模型的最新结果。
Flow models have recently made great progress at modeling ordinal discrete data such as images and audio. Due to the continuous nature of flow models, dequantization is typically applied when using them for such discrete data, resulting in lower bound estimates of the likelihood. In this paper, we introduce subset flows, a class of flows that can tractably transform finite volumes and thus allow exact computation of likelihoods for discrete data. Based on subset flows, we identify ordinal discrete autoregressive models, including WaveNets, PixelCNNs and Transformers, as single-layer flows. We use the flow formulation to compare models trained and evaluated with either the exact likelihood or its dequantization lower bound. Finally, we study multilayer flows composed of PixelCNNs and non-autoregressive coupling layers and demonstrate state-of-the-art results on CIFAR-10 for flow models trained with dequantization.
