SlideShare a Scribd company logo
1 of 19
1
DEEP LEARNING JP
[DL Papers]
http://deeplearning.jp/
Explorable Super-Resolution (CVPR2020)
Presenter: Kazutoshi Akita
(ToyotaTechnological Institute, IntelligentInformation Media Lab)
論文情報
• 論文名: Explorable Super Resolution (CVPR2020)
• 著者: Yuval Bahat, Tomer Michaeli
(Technion - Israel Institute of Technology, Haifa, Israel)
• URL
論文:
https://openaccess.thecvf.com/content_CVPR_2020/html/Bahat_Explorable_Super_Resolution_CVPR_2020_paper.html?fbclid=IwAR0Jws65Y7JQY
4L_bs67glpZcF17jeOQNiYNvUEVRVOhgkFHW4CDpg63yx8
プロジェクトページ:
https://yuvalbahat.github.io/Explorable-Super-Resolution/codes/
2
前提知識
• 超解像(Super Resolution, SR)
– 画像を綺麗に拡大する技術
3
超解像
(SRNet)
低解像画像
(LR画像)
超解像画像
(SR画像)
前提知識
• これまでの超解像
4
SRNet
低解像画像
(LR画像)
超解像画像
(SR画像)
・低解像化
・ブラー
高解像画像
(HR画像)
Loss (e.g., 二乗誤差,敵対的損失)
前提知識
• 異なる高解像画像が低解像化によって同じ画像になる = 不良設定問題
5
・低解像化
・ブラー
・低解像化
・ブラー
HR画像 LR画像
SRNet
SR画像
SRNet
NG
OK
前提知識
• 新タスク:出力をコントロールするシグナル𝑧により,様々なSR画像を生成
6
・低解像化
・ブラー
・低解像化
・ブラー
HR画像 LR画像
SRNet
SR画像
SRNet
OK
𝑧1
𝑧2
OK
論文概要
• Explorable Super Resolution
– 様々なSR画像を生成するという新たなタスクを開拓した論文の一つ
• SRFlow (ECCV 2020),DeepSEE (ACCV 2020) なども同様
– SR画像としての「制約」を担保するConsistency Enforcing Module (CEM)を提案
– ユーザーが直感的にコントロールシグナル𝑧を探索可能なGUIを構築
7
提案手法 (CEM)
• シグナル𝑧により,全く異なる画像が生成される可能性がある
• 制約:「出力SR画像を低解像化したら,入力LR画像と完全に一致する」
8
SRNet
入力LR画像
出力SR画像
・低解像化
・ブラー
一致
𝑧
提案手法 (CEM)
• Consistency Enforcing Module (CEM)
– SRの制約を満たしていないCNN出力(SR画像)を,制約を満たすように
修正するモジュール
9
提案手法 (CEM)
• 入力LR画像𝑦は,次のように表される
𝑥: 低解像化により同じ𝑦になりうるすべてのHR画像群, ℎ: ブラーカーネル
∗: 畳み込み, ↓α: 単純な補間方法などによる辺々
1
α
のダウンサンプリング処理
• ①は以下のように書き換え可能
10
𝑦 = (ℎ ∗ 𝑥) ↓α ・・・ ①
𝑦 = 𝐻𝑥 ・・・ ②
提案手法 (CEM)
• SRNetの出力群を𝑥𝑖𝑛𝑐とすると,以下の式を満たしたい
• 𝑥は,直交射影行列𝑃𝒩(𝐻)⊥, 𝑃𝒩(𝐻)を用いて,以下のように直交する2つのベクトルに分解可能
• ④を用いて③のminの中の式を変形すると,
11
min
𝑥
| 𝑥 − 𝑥𝑖𝑛𝑐 |2
𝐻𝑥 = 𝑦 ・・・ ③
s.t.
𝑃𝒩(𝐻)⊥ = 𝐻𝑇(𝐻𝐻𝑇)−1𝐻
𝑥 = 𝑃𝒩(𝐻)𝑥 + 𝑃𝒩(𝐻)⊥𝑥 ・・・ ④
𝑃𝒩 𝐻 = 𝐼 − 𝐻𝑇(𝐻𝐻𝑇)−1𝐻
||𝑃𝒩 𝐻 (𝑥 − 𝑥𝑖𝑛𝑐)||2 + ||𝑃𝒩(𝐻)⊥(𝑥 − 𝑥𝑖𝑛𝑐)||2 ・・・ ⑤
提案手法 (CEM)
• ⑤の第1項を0とする
– 分離した成分のうち片方では,𝑥と𝑥𝑖𝑛𝑐は完全に一致させることができると考える
– 一致させることができない要因(不良設定)はもう片方の成分に集約される
12
||𝑃𝒩 𝐻 (𝑥 − 𝑥𝑖𝑛𝑐)||2
+ ||𝑃𝒩(𝐻)⊥(𝑥 − 𝑥𝑖𝑛𝑐)||2
・・・ ⑤
𝑃𝒩 𝐻 𝑥 = 𝑃𝒩 𝐻 𝑥𝑖𝑛𝑐 ・・・ ⑥
提案手法 (CEM)
• これまでに得られた式を持って,𝑥を変形
13
𝑥 = 𝑃𝒩 𝐻 𝑥 + 𝑃𝒩(𝐻)⊥𝑥
= (𝐼 − 𝐻𝑇 𝐻𝐻𝑇 −1𝐻)𝑥 + 𝐻𝑇 𝐻𝐻𝑇 −1𝐻𝑥
= 𝐼 − 𝐻𝑇 𝐻𝐻𝑇 −1𝐻 𝑥 + 𝐻𝑇 𝐻𝐻𝑇 −1𝑦
= 𝑥𝑖𝑛𝑐 − ℎ ∗ 𝑘 ∗ ℎ ∗ 𝑥𝑖𝑛𝑐 ↓α ↑𝛼 +ℎ ∗ (𝑘 ∗ 𝑦) ↑𝛼
𝐻 = (ℎ ∗ ∙) ↓α
𝐻𝑇 = (ℎ ∗ ∙) ↑α
(𝐻𝐻𝑇
)−1
= 𝑘 ∗ ∙
提案手法 (CEM)
• 「出力SR画像を低解像化したら,入力LR画像と完全に一致する」制約を
かけていない従来のSR手法に組み込むことで,性能向上を確認
14
提案手法 (学習)
• 画像のエッジの強さ(=2階微分)に注目
– 画像の高周波な構造を最もよく表現している
15
ℒ𝑎𝑑𝑣 + λ𝑅𝑎𝑛𝑔𝑒ℒ𝑅𝑎𝑛𝑔𝑒
+λ𝑆𝑡𝑟𝑢𝑐𝑡ℒ𝑆𝑡𝑟𝑢𝑐𝑡 + λ𝑀𝑎𝑝ℒ𝑀𝑎𝑝
SRNet CEM
𝑧
LR画像
SR画像
提案手法 (学習)
• ℒ𝑆𝑡𝑟𝑢𝑐𝑡
– 𝑧で指定した通りのエッジがSR画像に生成されるように学習するLoss
• 入力の𝑧をランダムな空間的に一様な値とする
• SR画像を2階微分し,出力されたエッジと入力𝑧が一致するように学習
16
SRNet CEM
𝑧
2階微分
ℒ𝑆𝑡𝑟𝑢𝑐𝑡
LR画像
SR画像
提案手法 (学習)
• ℒ𝑀𝑎𝑝
– HR画像と一致する(=現実に観測される)SR画像を生成できるように学習するLoss
• 𝑧をランダムな空間的に一様な値で初期化し,SR画像とHR画像の差が最も小さい𝑧を探索
• ↑で見つけた𝑧におけるSR画像とHR画像の差をLossとする
17
SRNet CEM
𝑧
LR画像
SR画像
ℒ𝑀𝑎𝑝
HR画像
ℒ𝑀𝑎𝑝 = min
𝑧
| 𝜓 𝑦, 𝑧 − 𝑥|
𝜓(𝑦, 𝑧) 𝑥
提案手法 (直感的なSRの編集)
• 𝑧をユーザーが直感的に編集可能なGUIを構築
18
まとめ
• 超解像は,1つの低解像画像に対して複数の妥当な予測結果が存在する不良設定問題
• コントロールシグナルにより,様々なSR画像を出力できる確率的モデルを提案
• SRの制約である「SR画像を低解像化したら,入力LRに一致する」ことを必ず満たすよう
に出力の修正を行うCEMを提案
• 直感的に予測結果を編集できるフレームワークにより,ユーザーが望む超解像画像を
獲得可能
19

More Related Content

More from Deep Learning JP

More from Deep Learning JP (20)

【DL輪読会】Can Neural Network Memorization Be Localized?
【DL輪読会】Can Neural Network Memorization Be Localized?【DL輪読会】Can Neural Network Memorization Be Localized?
【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】Hopfield network 関連研究について【DL輪読会】Hopfield network 関連研究について
【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】SimPer: Simple self-supervised learning of periodic targets( ICLR 2023 )【DL輪読会】SimPer: Simple self-supervised learning of periodic targets( ICLR 2023 )
【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】RLCD: Reinforcement Learning from Contrast Distillation for Language M...【DL輪読会】RLCD: Reinforcement Learning from Contrast Distillation for Language M...
【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】"Secrets of RLHF in Large Language Models Part I: PPO"【DL輪読会】"Secrets of RLHF in Large Language Models Part I: PPO"
【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】"Language Instructed Reinforcement Learning  for Human-AI Coordination "【DL輪読会】"Language Instructed Reinforcement Learning  for Human-AI Coordination "
【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】Llama 2: Open Foundation and Fine-Tuned Chat Models【DL輪読会】Llama 2: Open Foundation and Fine-Tuned Chat Models
【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】"Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware"【DL輪読会】"Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware"
【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】Parameter is Not All You Need:Starting from Non-Parametric Networks fo...【DL輪読会】Parameter is Not All You Need:Starting from Non-Parametric Networks fo...
【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】Drag Your GAN: Interactive Point-based Manipulation on the Generative ...【DL輪読会】Drag Your GAN: Interactive Point-based Manipulation on the Generative ...
【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】Self-Supervised Learning from Images with a Joint-Embedding Predictive...【DL輪読会】Self-Supervised Learning from Images with a Joint-Embedding Predictive...
【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...
【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...
【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...
 
【DL輪読会】VIP: Towards Universal Visual Reward and Representation via Value-Impl...
【DL輪読会】VIP: Towards Universal Visual Reward and Representation via Value-Impl...【DL輪読会】VIP: Towards Universal Visual Reward and Representation via Value-Impl...
【DL輪読会】VIP: Towards Universal Visual Reward and Representation via Value-Impl...
 
【DL輪読会】Deep Transformers without Shortcuts: Modifying Self-attention for Fait...
【DL輪読会】Deep Transformers without Shortcuts: Modifying Self-attention for Fait...【DL輪読会】Deep Transformers without Shortcuts: Modifying Self-attention for Fait...
【DL輪読会】Deep Transformers without Shortcuts: Modifying Self-attention for Fait...
 
【DL輪読会】マルチモーダル 基盤モデル
【DL輪読会】マルチモーダル 基盤モデル【DL輪読会】マルチモーダル 基盤モデル
【DL輪読会】マルチモーダル 基盤モデル
 
【DL輪読会】TrOCR: Transformer-based Optical Character Recognition with Pre-traine...
【DL輪読会】TrOCR: Transformer-based Optical Character Recognition with Pre-traine...【DL輪読会】TrOCR: Transformer-based Optical Character Recognition with Pre-traine...
【DL輪読会】TrOCR: Transformer-based Optical Character Recognition with Pre-traine...
 
【DL輪読会】HyperDiffusion: Generating Implicit Neural Fields withWeight-Space Dif...
【DL輪読会】HyperDiffusion: Generating Implicit Neural Fields withWeight-Space Dif...【DL輪読会】HyperDiffusion: Generating Implicit Neural Fields withWeight-Space Dif...
【DL輪読会】HyperDiffusion: Generating Implicit Neural Fields withWeight-Space Dif...
 
【DL輪読会】大量API・ツールの扱いに特化したLLM
【DL輪読会】大量API・ツールの扱いに特化したLLM【DL輪読会】大量API・ツールの扱いに特化したLLM
【DL輪読会】大量API・ツールの扱いに特化したLLM
 
【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision
【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision
【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision
 
【DL輪読会】Poisoning Language Models During Instruction Tuning Instruction Tuning...
【DL輪読会】Poisoning Language Models During Instruction Tuning Instruction Tuning...【DL輪読会】Poisoning Language Models During Instruction Tuning Instruction Tuning...
【DL輪読会】Poisoning Language Models During Instruction Tuning Instruction Tuning...
 

Recently uploaded

Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire business
panagenda
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and Myths
Joaquim Jorge
 

Recently uploaded (20)

Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
GenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdfGenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdf
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire business
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and Myths
 
AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
Deploy with confidence: VMware Cloud Foundation 5.1 on next gen Dell PowerEdg...
Deploy with confidence: VMware Cloud Foundation 5.1 on next gen Dell PowerEdg...Deploy with confidence: VMware Cloud Foundation 5.1 on next gen Dell PowerEdg...
Deploy with confidence: VMware Cloud Foundation 5.1 on next gen Dell PowerEdg...
 
Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
 
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
 

【DL輪読会】Explorable Super Resolution