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
Signal Processing Course: Denoising
Search
Gabriel Peyré
January 01, 2012
Research
0
1.4k
Signal Processing Course: Denoising
Gabriel Peyré
January 01, 2012
Tweet
Share
More Decks by Gabriel Peyré
See All by Gabriel Peyré
The mathematics of transformers
gpeyre
0
88
Computational OT #4 - Gradient flow and diffusion models
gpeyre
0
830
Computational OT #1 - Monge and Kantorovitch
gpeyre
0
920
Computational OT #2 - Entropic regularization
gpeyre
0
440
Computational OT #3 - Dual and semidiscrete
gpeyre
0
290
Optimal and Diffusion Transports in Machine Learning
gpeyre
0
2.2k
The Mathematics of AI
gpeyre
1
2.4k
Transformers are Universal in Context Learners
gpeyre
0
980
Ground Metric Learning with applications in genomics
gpeyre
0
770
Other Decks in Research
See All in Research
svc-hook: hooking system calls on ARM64 by binary rewriting
retrage
2
140
さまざまなAgent FrameworkとAIエージェントの評価
ymd65536
1
440
LLMアプリケーションの透明性について
fufufukakaka
0
180
Community Driveプロジェクト(CDPJ)の中間報告
smartfukushilab1
0
200
LLM-jp-3 and beyond: Training Large Language Models
odashi
1
780
第二言語習得研究における 明示的・暗示的知識の再検討:この分類は何に役に立つか,何に役に立たないか
tam07pb915
0
1.4k
AIスーパーコンピュータにおけるLLM学習処理性能の計測と可観測性 / AI Supercomputer LLM Benchmarking and Observability
yuukit
1
710
音声感情認識技術の進展と展望
nagase
0
500
病院向け生成AIプロダクト開発の実践と課題
hagino3000
0
560
湯村研究室の紹介2025 / yumulab2025
yumulab
0
320
[Devfest Incheon 2025] 모두를 위한 친절한 언어모델(LLM) 학습 가이드
beomi
2
1.5k
英語教育 “研究” のあり方:学術知とアウトリーチの緊張関係
terasawat
1
410
Featured
See All Featured
Jess Joyce - The Pitfalls of Following Frameworks
techseoconnect
PRO
1
94
From Legacy to Launchpad: Building Startup-Ready Communities
dugsong
0
170
Fight the Zombie Pattern Library - RWD Summit 2016
marcelosomers
234
17k
The Impact of AI in SEO - AI Overviews June 2024 Edition
aleyda
5
760
So, you think you're a good person
axbom
PRO
2
1.9k
Data-driven link building: lessons from a $708K investment (BrightonSEO talk)
szymonslowik
1
950
Build your cross-platform service in a week with App Engine
jlugia
234
18k
We Are The Robots
honzajavorek
0
190
How to audit for AI Accessibility on your Front & Back End
davetheseo
0
210
Designing Experiences People Love
moore
143
24k
Designing for Timeless Needs
cassininazir
0
150
Rails Girls Zürich Keynote
gr2m
96
14k
Transcript
Linear and Non Linear Denoising Gabriel Peyré www.numerical-tours.com
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Noise in Images
Denoising Problem
Denoising Problem
Additive Noise Model
Noise Distributions −0.3 −0.2 −0.1 0 0.1 0.2 0.3 −0.3
−0.2 −0.1 0 0.1 0.2 0.3
Noise Distributions −0.3 −0.2 −0.1 0 0.1 0.2 0.3 −0.3
−0.2 −0.1 0 0.1 0.2 0.3
Noise Distributions −0.3 −0.2 −0.1 0 0.1 0.2 0.3 −0.3
−0.2 −0.1 0 0.1 0.2 0.3
Data-dependent Noise
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Linear Denoising Estimator
Fourier and Denoising
Optimal Filter Choice
Oracle Estimation of Optimal Filter
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Diagonal Thresholding
Wavelet Diagonal Hard Thresholding
Sparse Signal Estimation
Optimal Threshold Selection
W unit variance white noise. Non-linear Approximation and Estimation
Hard vs. Soft Thresholding
Hard vs. Soft Thresholding
Optimal Threshold
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Translation Invariant Denoising
Translation Invariant Wavelets
Translation Invariant Haar (1D)
Translation Invariant Transform (2D)
Translation Invariant Thresholding
Optimal Invariant Threshold
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Between Hard and Soft Thresholding
Stein Quadratic-Soft Thresholder
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Block Thresholding
Optimal Block Choice
Comparison
Overview • Noise in Signals and Images • Linear Denoising
by Blurring • Non-linear Wavelet Denoising • Translation Invariant Thresholding • Other Diagonal Thresholders • Non-diagonal Block Thresholding • Data-dependent Noise
Poisson Noise
Poisson Noise Variance Stabilization 1 2 3 4 5 6
7 8 9 10 0.75 0.8 0.85 0.9 0.95 1 1.05
Multiplicative Noise
Multiplicative Noise Stabilization 0 0.5 1 1.5 2 2.5 −1.5
−1 −0.5 0 0.5 1 1.5
Conclusion