Generating Sequences With Recurrent Neural Networks
Alex Graves Department of Computer Science University of Toronto
Abstract
This paper shows how Long Short-term Memory recurrent neural networks can be used to generate complex sequences with long-range structure, simply by predicting one data point at a time. The approach is demonstrated for text (where the data are discrete) and online handwriting (where the data are real-valued). It is then extended to handwriting synthesis by allowing the network to condition its predictions on a text sequence. The resulting system is able to generate highly realistic cursive handwriting in a wide variety of styles.
中文速览
长短期记忆网络（LSTM）天生擅长记住序列里的长程依赖关系，这篇论文正是利用这一特性，让网络每次只预测一个数据点，从而逐步"续写"出整条序列。作者先把这套方法用于文本生成（字符或词级别），再拓展到手写轨迹这类连续值数据，通过混合密度输出层让网络能够建模笔迹的概率分布；进一步地，他们引入了一个软注意力对齐机制，让网络在生成笔迹时能够"参照"输入的文字内容，从而实现手写合成（handwriting synthesis）。最终系统能够生成风格多样、高度逼真的草书手写体，部分样本肉眼难以与真实手写区分，还可以通过"风格引导"来模仿特定书写者的笔迹。这项工作证明了LSTM在连续序列生成上的强大潜力，为后来的序列生成模型和注意力机制研究奠定了重要基础。
原文 arXiv:1308.0850;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1308.0850v5