SlideShare a Scribd company logo
Censoring Representation
with Adversary
D3岩澤
書誌情報
• ICLR2016
• Harrison Edwards & Amos Storkey
• エディンバラ大学
• Citation:1
• 余談:5月にアップデートされて文がだいぶ読みやすくなって
る
概要
• GANで利用されているAdversarial Trainingを画像生成以外に
使う
• 具体的には,表現に含まれるべきでない情報を取り出す
• 含まれるべきでない例:
• Fairness:黒人/白人という情報で推薦結果を変えない
• Image Anonymization:画像中に含まれる住所や氏名を消したい
Fairness, Image Anonymizationの難しさ
• 例えば黒人/白人に依存しない推薦結果を与えたい場合を考える
• この時、特徴にある黒人/白人という特徴量を落とせば一見良さ
そうだがそれでは不十分
• なぜなら、他の特徴量が黒人、白人に関する情報量を持ってる
可能性があるから
• 素性が、黒人/白人に依存しないようにする必要がある
Fairness, Image Anonymizationの難しさ2
• データの価値とFairness/プライバシーは究極的に対立する
• 画像から機微情報を消したい場合
• 究極的には入力に依存しない特徴量を返す関数fを使えば機微情
報は消える
• がデータの価値も同時に損なわれる
• データの価値とプライバシー保護のバランスをとることが大事
提案手法
• 上を最適化
• 再構築誤差とクラス分類誤差は普通の誤差なので
Fairness/Privacyを説明する
• 以降Sとしては2値を考える
• 例:黒人か白人か、機微情報を含むか含まないか
再構築誤差 Fairness/Privacy クラス分類誤差
D(S,R)の学習: Adversarial Training
アルゴリズム全体
• 興味持った方は,結果などは論文を参照したください!
• https://arxiv.org/abs/1511.05897

More Related Content

PDF
[DL輪読会]Learning What and Where to Draw (NIPS’16)
PDF
[DL輪読会]最新の深層強化学習
PPTX
[DL輪読会]Exploiting Cyclic Symmetry in Convolutional Neural Networks
PDF
[DL輪読会]Wasserstein GAN/Towards Principled Methods for Training Generative Adv...
PPTX
[DL輪読会]Understanding deep learning requires rethinking generalization
PDF
[Dl輪読会]bridging the gaps between residual learning, recurrent neural networks...
PDF
[DL輪読会]Unsupervised Learning of 3D Structure from Images
PDF
[輪読会]Multilingual Image Description with Neural Sequence Models
[DL輪読会]Learning What and Where to Draw (NIPS’16)
[DL輪読会]最新の深層強化学習
[DL輪読会]Exploiting Cyclic Symmetry in Convolutional Neural Networks
[DL輪読会]Wasserstein GAN/Towards Principled Methods for Training Generative Adv...
[DL輪読会]Understanding deep learning requires rethinking generalization
[Dl輪読会]bridging the gaps between residual learning, recurrent neural networks...
[DL輪読会]Unsupervised Learning of 3D Structure from Images
[輪読会]Multilingual Image Description with Neural Sequence Models

Viewers also liked (16)

PPTX
[DL輪読会]Learning convolutional neural networks for graphs
PPTX
[DL輪読会]Let there be color
PPTX
[DL輪読会]TREE-STRUCTURED VARIATIONAL AUTOENCODER
PDF
[Dl輪読会]dl hacks輪読
PDF
[Dl輪読会]bayesian dark knowledge
PDF
Iclr2016 vaeまとめ
PDF
[DL輪読会]Combining Fully Convolutional and Recurrent Neural Networks for 3D Bio...
PPTX
[DL輪読会]Image-to-Image Translation with Conditional Adversarial Networks
PDF
[DL輪読会]Regularization with stochastic transformations and perturbations for d...
PDF
[DL輪読会]QUASI-RECURRENT NEURAL NETWORKS
PPTX
[DL輪読会]Unsupervised Cross-Domain Image Generation
PPTX
[DL輪読会]Wavenet a generative model for raw audio
PDF
[DL輪読会]StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generat...
PDF
[Dl輪読会]introduction of reinforcement learning
PDF
[DL輪読会]SeqGan Sequence Generative Adversarial Nets with Policy Gradient
PDF
FinalReport
[DL輪読会]Learning convolutional neural networks for graphs
[DL輪読会]Let there be color
[DL輪読会]TREE-STRUCTURED VARIATIONAL AUTOENCODER
[Dl輪読会]dl hacks輪読
[Dl輪読会]bayesian dark knowledge
Iclr2016 vaeまとめ
[DL輪読会]Combining Fully Convolutional and Recurrent Neural Networks for 3D Bio...
[DL輪読会]Image-to-Image Translation with Conditional Adversarial Networks
[DL輪読会]Regularization with stochastic transformations and perturbations for d...
[DL輪読会]QUASI-RECURRENT NEURAL NETWORKS
[DL輪読会]Unsupervised Cross-Domain Image Generation
[DL輪読会]Wavenet a generative model for raw audio
[DL輪読会]StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generat...
[Dl輪読会]introduction of reinforcement learning
[DL輪読会]SeqGan Sequence Generative Adversarial Nets with Policy Gradient
FinalReport
Ad

More from Deep Learning JP (20)

PPTX
【DL輪読会】AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners
PPTX
【DL輪読会】事前学習用データセットについて
PPTX
【DL輪読会】 "Learning to render novel views from wide-baseline stereo pairs." CVP...
PPTX
【DL輪読会】Zero-Shot Dual-Lens Super-Resolution
PPTX
【DL輪読会】BloombergGPT: A Large Language Model for Finance arxiv
PPTX
【DL輪読会】マルチモーダル LLM
PDF
【 DL輪読会】ToolLLM: Facilitating Large Language Models to Master 16000+ Real-wo...
PPTX
【DL輪読会】AnyLoc: Towards Universal Visual Place Recognition
PDF
【DL輪読会】Can Neural Network Memorization Be Localized?
PPTX
【DL輪読会】Hopfield network 関連研究について
PPTX
【DL輪読会】SimPer: Simple self-supervised learning of periodic targets( ICLR 2023 )
PDF
【DL輪読会】RLCD: Reinforcement Learning from Contrast Distillation for Language M...
PDF
【DL輪読会】"Secrets of RLHF in Large Language Models Part I: PPO"
PPTX
【DL輪読会】"Language Instructed Reinforcement Learning for Human-AI Coordination "
PPTX
【DL輪読会】Llama 2: Open Foundation and Fine-Tuned Chat Models
PDF
【DL輪読会】"Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware"
PPTX
【DL輪読会】Parameter is Not All You Need:Starting from Non-Parametric Networks fo...
PDF
【DL輪読会】Drag Your GAN: Interactive Point-based Manipulation on the Generative ...
PDF
【DL輪読会】Self-Supervised Learning from Images with a Joint-Embedding Predictive...
PPTX
【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...
【DL輪読会】AdaptDiffuser: Diffusion Models as Adaptive Self-evolving Planners
【DL輪読会】事前学習用データセットについて
【DL輪読会】 "Learning to render novel views from wide-baseline stereo pairs." CVP...
【DL輪読会】Zero-Shot Dual-Lens Super-Resolution
【DL輪読会】BloombergGPT: A Large Language Model for Finance arxiv
【DL輪読会】マルチモーダル LLM
【 DL輪読会】ToolLLM: Facilitating Large Language Models to Master 16000+ Real-wo...
【DL輪読会】AnyLoc: Towards Universal Visual Place Recognition
【DL輪読会】Can Neural Network Memorization Be Localized?
【DL輪読会】Hopfield network 関連研究について
【DL輪読会】SimPer: Simple self-supervised learning of periodic targets( ICLR 2023 )
【DL輪読会】RLCD: Reinforcement Learning from Contrast Distillation for Language M...
【DL輪読会】"Secrets of RLHF in Large Language Models Part I: PPO"
【DL輪読会】"Language Instructed Reinforcement Learning for Human-AI Coordination "
【DL輪読会】Llama 2: Open Foundation and Fine-Tuned Chat Models
【DL輪読会】"Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware"
【DL輪読会】Parameter is Not All You Need:Starting from Non-Parametric Networks fo...
【DL輪読会】Drag Your GAN: Interactive Point-based Manipulation on the Generative ...
【DL輪読会】Self-Supervised Learning from Images with a Joint-Embedding Predictive...
【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...
Ad

[Dl輪読会]Censoring Representation with Adversary