Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
MixPoet
Search
Zhang Yixiao
April 30, 2020
Research
4
380
MixPoet
Zhang Yixiao
April 30, 2020
Tweet
Share
More Decks by Zhang Yixiao
See All by Zhang Yixiao
CoCon
ldzhangyx
0
350
vq-cpc
ldzhangyx
0
350
diora
ldzhangyx
0
250
drummernet
ldzhangyx
0
200
ON-LSTM
ldzhangyx
0
160
Other Decks in Research
See All in Research
プロシェアリング白書2025_PROSHARING_REPORT_2025
circulation
1
700
Large Language Model Agent: A Survey on Methodology, Applications and Challenges
shunk031
11
6.7k
Transparency to sustain open science infrastructure - Printemps Couperin
mlarrieu
1
140
ストレス計測方法の確立に向けたマルチモーダルデータの活用
yurikomium
0
260
ウッドスタックチャン:木材を用いた小型エージェントロボットの開発と印象評価 / ec75-sato
yumulab
1
330
2025年度人工知能学会全国大会チュートリアル講演「深層基盤モデルの数理」
taiji_suzuki
22
11k
数理最適化に基づく制御
mickey_kubo
5
630
Agentic AIとMCPを利用したサービス作成入門
mickey_kubo
0
130
SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery
satai
3
160
ノンパラメトリック分布表現を用いた位置尤度場周辺化によるRTK-GNSSの整数アンビギュイティ推定
aoki_nosse
0
310
AIによる画像認識技術の進化 -25年の技術変遷を振り返る-
hf149
6
2.7k
電力システム最適化入門
mickey_kubo
1
540
Featured
See All Featured
Embracing the Ebb and Flow
colly
85
4.7k
Why You Should Never Use an ORM
jnunemaker
PRO
56
9.4k
Rebuilding a faster, lazier Slack
samanthasiow
81
9k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
367
26k
Stop Working from a Prison Cell
hatefulcrawdad
269
20k
Build your cross-platform service in a week with App Engine
jlugia
231
18k
XXLCSS - How to scale CSS and keep your sanity
sugarenia
248
1.3M
The Power of CSS Pseudo Elements
geoffreycrofte
76
5.8k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
228
22k
Statistics for Hackers
jakevdp
799
220k
YesSQL, Process and Tooling at Scale
rocio
172
14k
The Pragmatic Product Professional
lauravandoore
35
6.7k
Transcript
MixPoet: Diverse Poetry Generation via Learning Controllable Mixed Latent Space
ArXiv: 2003.06094v1 Presenter: Yixiao Zhang
Overview • Idea: 诗人经历、历史背景等 => 诗歌风格多样化 • Methods: • semi-supervised
VAE • disentangling latent space to sub-spaces • each sub-space corresponds to one factor conditioning • adversarial training
Introduction • 近年的研究,主要考虑语义连贯、主题相关 • 存在diversity的困扰 • diversity: • 主题间多样性:给定两个topic words,生成不同的诗歌
• 主题内多样性:给定一个topic word,生成不同的诗歌 • * 现有的模型倾向于记住常见pattern
Introduction • 生活经历、历史背景、文学流派 => 影响风格
Introduction • MixPoet: semi-supervised VAE • 将latent space分解为sub-spaces,与影响因子一一对应 • 训练阶段:模型预测无label诗歌的factors
• 测试阶段:指定factor的值,生成风格化的诗歌
Related Work • 诗歌生成模型 (RNNs, Memory Models, etc. ) •
多样性的先前研究: • MRL system: 强化学习,鼓励选用高TF-IDF的词汇 • USPG: 无监督最大化style vector和诗歌的mutual information
Related Work • VAE文本生成/诗歌生成 • Yang et. al, 2018b: 学习context-conditioned
latent variable • Hu et al. 2017: 对生成的诗歌进行对抗训练,增强topic相关性 • CVAE 对话多样性: Learning Discourse-level Diversity for Neural Dialog Models using Conditional Variational Autoencoders, ACL 2017 • 本文的对抗:在latent space上做对抗训练
Method • topic keyword: mixture empirical distributions: labeled/ unlabeled
Method: Generator • GRU based model • 是length embedding
Method: Semi-supervised C-VAE • 目的是学习 • 引入z • 由于style与semantics耦合 •
不假设y与z的独立性,而是: • 顺序: w => y => z => x (无y label时)
Method: Semi-supervised C-VAE • then for labeled data: • 估计先验
• 和后验 分别使用一个network计算, recon时最小化KL散度。
Method: Semi-supervised C-VAE • labeled data is too limited •
将y看作另一个latent variable • 估计先验 • 和后验 分别使用一个MLP network计算, recon y时最小化KL散度。
Method: Semi-supervised C-VAE • Total Loss:
Method: Latent Space Mixture • 多个factor时的情形: • 独立性假设:
Method: Latent Space Mixture • How to learn mixed latent
space? • For Isotropic Gaussian Space:
Method: Latent Space Mixture • How to learn mixed latent
space? • For Universal Space: 对于condition: ita是噪声,delta是脉冲函数,c是w, y => 从分布中sample出一个值
Method: Latent Space Mixture • 之后使得discriminator区分这两个z • 估计KL散度: • 其中
就是discriminator
Experiments • factors: • 军旅生涯, 乡村生活, 其他 • 时代繁荣, 时代衰落
• => 6种style
Experiments • Baseline: • Ground Truth • C-VAE • USPG
• MRL: SOTA • fBasic, 监督学习模型
Experiments • 多样性,使用Jaccard Similarity指数评价,越低越好 • 诗歌质量:使用Language Model Score(LMS)评价 • 观察:
• 大多数模型倾向生成重复的短语 • MRL与Basic在intra部分只能生成极其相似的诗歌 • C-VAE情况类似
Experiments • Factor Control Results: • 测试生成的诗歌是否与给定因子类别一致
Experiments • 主观实验
Analysis: Style Mixture
Analysis