SlideShare une entreprise Scribd logo
1  sur  47
Compressed Sensing:
Introduction and Apps
Achuta Kadambi
Camera Culture, MIT
Exploiting Signals
• Not all signals are equal! Find a weakness then exploit.
Exploiting Signals
• Not all signals are equal! Find a weakness then exploit.

• Shannon-Nyquist  Bandlimited signals can be
sampled/reconstructed
Exploiting Signals
• Not all signals are equal! Find a weakness then exploit.

• Shannon-Nyquist  Bandlimited signals can be
sampled/reconstructed
• Rank-constrained Optimization  Low Rank signals can be
interpolated (Netflix Problem)
Exploiting Signals
• Not all signals are equal! Find a weakness then exploit.

• Shannon-Nyquist  Bandlimited signals can be
sampled/reconstructed
• Rank-constrained Optimization  Low Rank signals can be
interpolated. (Netflix Problem)
• Compressed Sensing  Sparse signals can be undersampled and
recovered.
Outline of this talk.
• Compressed Sensing overview.

• Very brief explanation on the why and how of Compressed Sensing.
• ‘Apps’ that use compressed sensing.
• Practical strategies for implementation (e.g. pseudocode, libraries).
Motivation: JPEG Compression
Our visual system is less sensitive to high
(spatial) frequency detail. Can we throw away
these frequencies and retain a similar image?

This is the intuition behind JPEG.

Spatial Frequency E.g.:
High  Hair, Blades of Grass, etc.
Low  Sky, Skin, etc.

Compressed Sensing: If we are going to throw
away stuff … why spend time acquiring it?
E.g.
Wired Magazine: “Fill in the Blanks…”
1D Implementation in L1Magic
Step 1: The original signal and its Fourier Transform.

Original Signal (N = 256)

Spectrum
Implementation in L1Magic
Step 2: The subsampled signal
Red Entries (80 samples) are observed.

Blue Entries (176 samples) must be
recovered.

That means we observe only 30% of the
original signal.
Implementation in L1Magic
Step 3: Exact Recovery of the Signal.

Original Signal (N = 256)

Reconstruction (N = 256)
L1Magic for Images

Original Image:
1 million pixels

Reconstruction:
from 100,000
random measurements.
Goes back to Fourier
Fourier Transform

Intuition: Projection, or Inner Product, of Signal with Trigonometric Functions.
Sparsity goes back to Fourier (circa
1800)
Superposition of Sinusoids

Original Time Domain Function

Frequency Domain Representation
Discrete Fourier Transform

Example DFT:
Time Signal is a Delta. Spectrum is Broadband.
DFT in Matrix Form
Nyquist-Shannon Sampling Theorem
In Shannon’s words:

How to Reconstruct? (Interpolation)

Compressive Sensing: Can we do better?
Inverse Problem
Example 1: Sinc Interpolation. Given the Data (a sufficiently sampled
signal), how can we obtain the original signal?

Example 2: Blurry Photos. Given a Blurry Photo, from a Camera, how can
we go back to the original, sharp image?
Example 3: Given a discrete time signal, how can we obtain its discrete
spectrum? **DFT problem is a Linear Inverse Problem
Solving the DFT Problem
Done?
Solving the DFT Problem via
Optimization
Done?

Loss Function
Solving the DFT Problem via
Optimization
Done?

PseudoInverse:
Minimize MSE
Constraining our Solution via
Regularization
We can go beyond loss function,
e.g., Tikhonov Regularization

Additional Term allows for some prior on original signal. For
instance if Tikhonov Matrix is a first order difference, then you
are biasing x toward smooth solutions.
Linked to the Lagrange problem, as well as Maximum A
Posteriori from probability, and Weiner filter from Sig proc.
Compressed Sensing
Compressed Sensing Structure
• Underdetermined system. y=Ax.
•
•
•
•

Y is m-dimensional sampled vector
A is mxn matrix
X is n-dimensional original vector.
And m << n

y

A

x
Simply Solving y=Ax not good enough
• This gives you an affine space with many solutions to y=Ax.

• So we must constrain our problem to look for the sparse solution to
y=Ax.
Occams Razor
Occam's Razor: among otherwise equal explanations,
the simplest is best
Occams Razor
Occam's Razor: among otherwise equal explanations,
the simplest is best

CS Occam's Razor: among otherwise equal solutions, the sparsest is best

Unfortunately, this optimization is not tractable
Geometric Property of Norms
The l1 Optimization Problem
RIP/Spark/Coherence
• The sensing matrix A must be carefully chosen.

• For compressed sensing to work, the matrix A must satisfy the
Restricted Isometry Property (RIP):

• Calculating RIP is NP-hard. We can work with easier quantities than
the RIP, such as spark and mutual incoherence.
App1: Single-Pixel Camera
App1: Single-Pixel Camera

Design
Advantage: A
MP camera
with just a
single-pixel.
App2: Single-Pixel THz imaging.
Design Advantage: CS allows for
single-pixel THz sensors, which are
much easier to fabricate than pixel
array.

In general, you can buy amazing
things at single-pixel level, e.g.,
picosecond detectors, thermal IR
sensor, etc.

Chan et al. Applied Physics
2008
App2: Monitoring Breathing via
Smartphone
Very similar to
OMP, DFT
formulation.
Basically finding
sparse spectral
components that
characterize the
audio signal of
breathing.

Design Advantage: Compressed sensing
allows for low power acquisition and
reduced streaming.

Oletic, Skrapec, and Bilas MobiHealth 2012
App4: Biometrics … Face Recognition
Design Advantage: Using
compressed sensing to handle
the small sample size problem.
Before, the number of samples in
the database Is less than the
degrees of freedom of each
sample.
App5: Fast MRI

Design Advantage:
Less samples means
less time for an MRI
Scan, which means
less time a sick or
disabled patient lies
in the scanner.
App5: Fast MRI
App6: Compressive Sensing of High
Speed Periodic Videos
Design Advantage: Exploit
sparsity of Periodic Videos to
obtain a high speed video
without using a high speed
camera.

Veeraraghavan A, Reddy
D, Raskar R. IEEE PAMI
App7: Compressive Light Field
Photography
Marwah, Wetzstein, Bando, Ra
skar. ACM SIGGRAPH 2013.
Design Advantage:
Obtain High-Resolution
Light Field photos by
placing a coded mask in
front of the sensor.
App8: Sparsity-Induced Time of Flight
Cameras

Kadambi et al. ACM SIGGRAPH Asia

E.g., Light Sweep Movies from Refael’s talk last week. Goal is to
obtain a well-conditioned deconvolution problem.

Design Advantage: Deconvolve to obtain
bounces of light and construct a light
sweep video.
App9  YOUR App!
Design Advantage: <insert here>
Practical Strategies
Many libraries are available for C++/Matlab/etc.
• Recommended: L1magic (http://users.ece.gatech.edu/~justin/l1magic/)
• SPGL1 (http://www.cs.ubc.ca/~mpf/spgl1/)

• CVX (http://cvxr.com/cvx/)
• On phone, nothing exists yet, but you can use Efficient Java Matrix Library
(EJML) to implement solely in linear algebra.
L1Magic Pseudocode
For 1D signal.
x = original_signal;
R = randn(m,n);
A = orth(R’)’;

// read in your original signal.
// create a random matrix of dimension mxn
// sensing matrix with orthogonal columns.

y = A*x;
// create subsampled signal y of only m entries.
X0 = A’*y;
// Initial guess by taking matrix inverse.
x_hat = l1eq_pd(x0, A, [], y, 1e-3); // run l1 solver.
norm(x_hat – x);

// error; should be zero in ideal case
Take-home Messages
• Opportunity to integrate cutting-edge mathematical techniques into
your camera apps.
• Compressed Sensing is lightweight in terms of coding. The key is
correctly identifying the sparsity in your engineering problem.
• Not all signals are equal… find a weakness, e.g., sparsity/rank, and
exploit it.
• Exploit in Hardware, Exploit in Software…

Contenu connexe

Tendances

Color image processing Presentation
Color image processing PresentationColor image processing Presentation
Color image processing PresentationRevanth Chimmani
 
Fir filter design using windows
Fir filter design using windowsFir filter design using windows
Fir filter design using windowsSarang Joshi
 
M ary psk modulation
M ary psk modulationM ary psk modulation
M ary psk modulationAhmed Diaa
 
Mimo in Wireless Communication
Mimo in Wireless CommunicationMimo in Wireless Communication
Mimo in Wireless Communicationkailash karki
 
Image Enhancement - Point Processing
Image Enhancement - Point ProcessingImage Enhancement - Point Processing
Image Enhancement - Point ProcessingGayathri31093
 
Image trnsformations
Image trnsformationsImage trnsformations
Image trnsformationsJohn Williams
 
Digital speech processing lecture1
Digital speech processing lecture1Digital speech processing lecture1
Digital speech processing lecture1Samiul Parag
 
4.4 diversity combining techniques
4.4   diversity combining techniques4.4   diversity combining techniques
4.4 diversity combining techniquesJAIGANESH SEKAR
 
Spatial filtering using image processing
Spatial filtering using image processingSpatial filtering using image processing
Spatial filtering using image processingAnuj Arora
 
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...IRJET Journal
 
Watershed Segmentation Image Processing
Watershed Segmentation Image ProcessingWatershed Segmentation Image Processing
Watershed Segmentation Image ProcessingArshad Hussain
 
(Full MatLab Code) Image compression DCT
(Full MatLab Code) Image compression DCT(Full MatLab Code) Image compression DCT
(Full MatLab Code) Image compression DCTChaudhary Sarimurrab
 
The evolution of TMS, family of DSP\'s
The evolution of TMS, family of DSP\'sThe evolution of TMS, family of DSP\'s
The evolution of TMS, family of DSP\'sRitul Sonania
 
06 spatial filtering DIP
06 spatial filtering DIP06 spatial filtering DIP
06 spatial filtering DIPbabak danyal
 

Tendances (20)

Source coding
Source coding Source coding
Source coding
 
SPACE DIVISION MULTIPLEXING (SDMA)
SPACE DIVISION MULTIPLEXING (SDMA)SPACE DIVISION MULTIPLEXING (SDMA)
SPACE DIVISION MULTIPLEXING (SDMA)
 
Color image processing Presentation
Color image processing PresentationColor image processing Presentation
Color image processing Presentation
 
Fir filter design using windows
Fir filter design using windowsFir filter design using windows
Fir filter design using windows
 
M ary psk modulation
M ary psk modulationM ary psk modulation
M ary psk modulation
 
Mimo in Wireless Communication
Mimo in Wireless CommunicationMimo in Wireless Communication
Mimo in Wireless Communication
 
Image Enhancement - Point Processing
Image Enhancement - Point ProcessingImage Enhancement - Point Processing
Image Enhancement - Point Processing
 
Image trnsformations
Image trnsformationsImage trnsformations
Image trnsformations
 
Digital speech processing lecture1
Digital speech processing lecture1Digital speech processing lecture1
Digital speech processing lecture1
 
4.4 diversity combining techniques
4.4   diversity combining techniques4.4   diversity combining techniques
4.4 diversity combining techniques
 
Spatial filtering using image processing
Spatial filtering using image processingSpatial filtering using image processing
Spatial filtering using image processing
 
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...
Implementation of FPGA Based Image Processing Algorithm using Xilinx System G...
 
Watershed Segmentation Image Processing
Watershed Segmentation Image ProcessingWatershed Segmentation Image Processing
Watershed Segmentation Image Processing
 
(Full MatLab Code) Image compression DCT
(Full MatLab Code) Image compression DCT(Full MatLab Code) Image compression DCT
(Full MatLab Code) Image compression DCT
 
The evolution of TMS, family of DSP\'s
The evolution of TMS, family of DSP\'sThe evolution of TMS, family of DSP\'s
The evolution of TMS, family of DSP\'s
 
06 spatial filtering DIP
06 spatial filtering DIP06 spatial filtering DIP
06 spatial filtering DIP
 
Spiral Antenna
Spiral Antenna  Spiral Antenna
Spiral Antenna
 
rake reciever ppt
rake reciever pptrake reciever ppt
rake reciever ppt
 
Wiener Filter
Wiener FilterWiener Filter
Wiener Filter
 
Bit error rate
Bit error rateBit error rate
Bit error rate
 

En vedette

Introduction to compressive sensing
Introduction to compressive sensingIntroduction to compressive sensing
Introduction to compressive sensingMohammed Musfir N N
 
Compressive Sensing Basics - Medical Imaging - MRI
Compressive Sensing Basics - Medical Imaging - MRICompressive Sensing Basics - Medical Imaging - MRI
Compressive Sensing Basics - Medical Imaging - MRIThomas Stefani
 
Nonconvex Compressed Sensing with the Sum-of-Squares Method
Nonconvex Compressed Sensing with the Sum-of-Squares MethodNonconvex Compressed Sensing with the Sum-of-Squares Method
Nonconvex Compressed Sensing with the Sum-of-Squares MethodTasuku Soma
 
Recovering Lost Sensor Data through Compressed Sensing
Recovering Lost Sensor Data through Compressed SensingRecovering Lost Sensor Data through Compressed Sensing
Recovering Lost Sensor Data through Compressed SensingZainul Charbiwala
 

En vedette (20)

Introduction to compressive sensing
Introduction to compressive sensingIntroduction to compressive sensing
Introduction to compressive sensing
 
Compressive Sensing Basics - Medical Imaging - MRI
Compressive Sensing Basics - Medical Imaging - MRICompressive Sensing Basics - Medical Imaging - MRI
Compressive Sensing Basics - Medical Imaging - MRI
 
Nonconvex Compressed Sensing with the Sum-of-Squares Method
Nonconvex Compressed Sensing with the Sum-of-Squares MethodNonconvex Compressed Sensing with the Sum-of-Squares Method
Nonconvex Compressed Sensing with the Sum-of-Squares Method
 
Multiview Imaging HW Overview
Multiview Imaging HW OverviewMultiview Imaging HW Overview
Multiview Imaging HW Overview
 
Coded Photography - Ramesh Raskar
Coded Photography - Ramesh RaskarCoded Photography - Ramesh Raskar
Coded Photography - Ramesh Raskar
 
Stereo and 3D Displays - Matt Hirsch
Stereo and 3D Displays - Matt HirschStereo and 3D Displays - Matt Hirsch
Stereo and 3D Displays - Matt Hirsch
 
What is Media in MIT Media Lab, Why 'Camera Culture'
What is Media in MIT Media Lab, Why 'Camera Culture'What is Media in MIT Media Lab, Why 'Camera Culture'
What is Media in MIT Media Lab, Why 'Camera Culture'
 
Google Glass Breakdown
Google Glass BreakdownGoogle Glass Breakdown
Google Glass Breakdown
 
What is SIGGRAPH NEXT? Intro by Ramesh Raskar
What is SIGGRAPH NEXT? Intro by Ramesh RaskarWhat is SIGGRAPH NEXT? Intro by Ramesh Raskar
What is SIGGRAPH NEXT? Intro by Ramesh Raskar
 
Raskar UIST Keynote 2015 November
Raskar UIST Keynote 2015 NovemberRaskar UIST Keynote 2015 November
Raskar UIST Keynote 2015 November
 
Leap Motion Development (Rohan Puri)
Leap Motion Development (Rohan Puri)Leap Motion Development (Rohan Puri)
Leap Motion Development (Rohan Puri)
 
Introduction to Camera Challenges - Ramesh Raskar
Introduction to Camera Challenges - Ramesh RaskarIntroduction to Camera Challenges - Ramesh Raskar
Introduction to Camera Challenges - Ramesh Raskar
 
Google Glass Overview
Google Glass OverviewGoogle Glass Overview
Google Glass Overview
 
Raskar stanfordextremecompuimagingapr2016
Raskar stanfordextremecompuimagingapr2016Raskar stanfordextremecompuimagingapr2016
Raskar stanfordextremecompuimagingapr2016
 
Kinect Tutorial
Kinect Tutorial Kinect Tutorial
Kinect Tutorial
 
Light Field Photography Introduction
Light Field Photography IntroductionLight Field Photography Introduction
Light Field Photography Introduction
 
Introduction to Photography
Introduction to PhotographyIntroduction to Photography
Introduction to Photography
 
Time of Flight Cameras - Refael Whyte
Time of Flight Cameras - Refael WhyteTime of Flight Cameras - Refael Whyte
Time of Flight Cameras - Refael Whyte
 
Recovering Lost Sensor Data through Compressed Sensing
Recovering Lost Sensor Data through Compressed SensingRecovering Lost Sensor Data through Compressed Sensing
Recovering Lost Sensor Data through Compressed Sensing
 
Developing "True HDR" for the iPhone
Developing "True HDR" for the iPhoneDeveloping "True HDR" for the iPhone
Developing "True HDR" for the iPhone
 

Similaire à Compressed Sensing - Achuta Kadambi

>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...
>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...
>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...Matt Hirsch - MIT Media Lab
 
Compressive Sampling Presentation
Compressive Sampling PresentationCompressive Sampling Presentation
Compressive Sampling PresentationManu Mitra
 
5.2. lithography 3,4,5 final,2013
5.2. lithography 3,4,5 final,20135.2. lithography 3,4,5 final,2013
5.2. lithography 3,4,5 final,2013Bhargav Veepuri
 
Ee 417 Senior Design
Ee 417 Senior DesignEe 417 Senior Design
Ee 417 Senior Designcrouchj1
 
Signal proccessing(wavelet) on radio telescopic images
Signal proccessing(wavelet) on radio telescopic imagesSignal proccessing(wavelet) on radio telescopic images
Signal proccessing(wavelet) on radio telescopic imagesshantanu Chutiya begger
 
A computer vision approach to speech enhancement
A computer vision approach to speech enhancementA computer vision approach to speech enhancement
A computer vision approach to speech enhancementRamin Anushiravani
 
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...KAIST
 
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...KAIST
 
Digital Theory 1.pdf
Digital Theory 1.pdfDigital Theory 1.pdf
Digital Theory 1.pdfssuserccc5db
 
Automated Speech Recognition
Automated Speech Recognition Automated Speech Recognition
Automated Speech Recognition Pruthvij Thakar
 
Volumetric Lighting for Many Lights in Lords of the Fallen
Volumetric Lighting for Many Lights in Lords of the FallenVolumetric Lighting for Many Lights in Lords of the Fallen
Volumetric Lighting for Many Lights in Lords of the FallenBenjamin Glatzel
 
High-Speed Single-Photon SPAD Camera
High-Speed Single-Photon SPAD CameraHigh-Speed Single-Photon SPAD Camera
High-Speed Single-Photon SPAD CameraFabrizio Guerrieri
 
Acoustic echo cancellation
Acoustic echo cancellationAcoustic echo cancellation
Acoustic echo cancellationchintanajoshi
 
Deep learning from a novice perspective
Deep learning from a novice perspectiveDeep learning from a novice perspective
Deep learning from a novice perspectiveAnirban Santara
 
Strength of Materials iLab
Strength of Materials iLabStrength of Materials iLab
Strength of Materials iLabBabatunde Ishola
 
CR (Computed Radiography) System
CR (Computed Radiography) SystemCR (Computed Radiography) System
CR (Computed Radiography) SystemJNDRSHD
 

Similaire à Compressed Sensing - Achuta Kadambi (20)

>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...
>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...
>A Switchable Light Field Camera Architecture with Angle SEnsitive Pixels and...
 
Lightspeed SIGGRAPH talk
Lightspeed SIGGRAPH talkLightspeed SIGGRAPH talk
Lightspeed SIGGRAPH talk
 
Compressive Sampling Presentation
Compressive Sampling PresentationCompressive Sampling Presentation
Compressive Sampling Presentation
 
5.2. lithography 3,4,5 final,2013
5.2. lithography 3,4,5 final,20135.2. lithography 3,4,5 final,2013
5.2. lithography 3,4,5 final,2013
 
Sparse and Redundant Representations: Theory and Applications
Sparse and Redundant Representations: Theory and ApplicationsSparse and Redundant Representations: Theory and Applications
Sparse and Redundant Representations: Theory and Applications
 
Ee 417 Senior Design
Ee 417 Senior DesignEe 417 Senior Design
Ee 417 Senior Design
 
Signal proccessing(wavelet) on radio telescopic images
Signal proccessing(wavelet) on radio telescopic imagesSignal proccessing(wavelet) on radio telescopic images
Signal proccessing(wavelet) on radio telescopic images
 
A computer vision approach to speech enhancement
A computer vision approach to speech enhancementA computer vision approach to speech enhancement
A computer vision approach to speech enhancement
 
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
 
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
[AAAI2018] Multispectral Transfer Network: Unsupervised Depth Estimation for ...
 
Digital Theory 1.pdf
Digital Theory 1.pdfDigital Theory 1.pdf
Digital Theory 1.pdf
 
Automated Speech Recognition
Automated Speech Recognition Automated Speech Recognition
Automated Speech Recognition
 
Volumetric Lighting for Many Lights in Lords of the Fallen
Volumetric Lighting for Many Lights in Lords of the FallenVolumetric Lighting for Many Lights in Lords of the Fallen
Volumetric Lighting for Many Lights in Lords of the Fallen
 
High-Speed Single-Photon SPAD Camera
High-Speed Single-Photon SPAD CameraHigh-Speed Single-Photon SPAD Camera
High-Speed Single-Photon SPAD Camera
 
Acoustic echo cancellation
Acoustic echo cancellationAcoustic echo cancellation
Acoustic echo cancellation
 
Computational Photography_TED.pptx
Computational Photography_TED.pptxComputational Photography_TED.pptx
Computational Photography_TED.pptx
 
FermiPoster
FermiPosterFermiPoster
FermiPoster
 
Deep learning from a novice perspective
Deep learning from a novice perspectiveDeep learning from a novice perspective
Deep learning from a novice perspective
 
Strength of Materials iLab
Strength of Materials iLabStrength of Materials iLab
Strength of Materials iLab
 
CR (Computed Radiography) System
CR (Computed Radiography) SystemCR (Computed Radiography) System
CR (Computed Radiography) System
 

Plus de Camera Culture Group, MIT Media Lab

God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar
God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar
God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar Camera Culture Group, MIT Media Lab
 
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...Dont follow the rainbow: How to avoid career traps that can lead you to fail,...
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...Camera Culture Group, MIT Media Lab
 
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019Camera Culture Group, MIT Media Lab
 
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...Camera Culture Group, MIT Media Lab
 

Plus de Camera Culture Group, MIT Media Lab (14)

Raskar Sig2017 Siggraph Achievement Award Talk
Raskar Sig2017 Siggraph Achievement Award TalkRaskar Sig2017 Siggraph Achievement Award Talk
Raskar Sig2017 Siggraph Achievement Award Talk
 
Lost Decade of Computational Photography
Lost Decade of Computational PhotographyLost Decade of Computational Photography
Lost Decade of Computational Photography
 
Covid Safe Paths
Covid Safe PathsCovid Safe Paths
Covid Safe Paths
 
God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar
God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar
God’s Eye View: Will global AI empower us or destroy us? | Ramesh Raskar
 
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...Dont follow the rainbow: How to avoid career traps that can lead you to fail,...
Dont follow the rainbow: How to avoid career traps that can lead you to fail,...
 
Raskar PhD and MS Thesis Guidance
Raskar PhD and MS Thesis GuidanceRaskar PhD and MS Thesis Guidance
Raskar PhD and MS Thesis Guidance
 
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019
Making Invisible Visible, Ramesh Raskar Keynote at Embedded Vision 2019
 
Augmented Surgeons: AI AR for Anatome, Raskar Aria 2019
Augmented Surgeons: AI AR for Anatome, Raskar Aria 2019Augmented Surgeons: AI AR for Anatome, Raskar Aria 2019
Augmented Surgeons: AI AR for Anatome, Raskar Aria 2019
 
Geo-spatial Research: Transition from Analysis to Synthesis
Geo-spatial Research: Transition from Analysis to SynthesisGeo-spatial Research: Transition from Analysis to Synthesis
Geo-spatial Research: Transition from Analysis to Synthesis
 
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...
Split Learning versus Federated Learning for Data Transparent ML, Camera Cult...
 
Unspoken Challenges in AR and XR
Unspoken Challenges in AR and XRUnspoken Challenges in AR and XR
Unspoken Challenges in AR and XR
 
Computer Vision Introduction
Computer Vision IntroductionComputer Vision Introduction
Computer Vision Introduction
 
Introduction to Light Fields
Introduction to Light FieldsIntroduction to Light Fields
Introduction to Light Fields
 
Raskar TEDMED 2013
Raskar TEDMED 2013Raskar TEDMED 2013
Raskar TEDMED 2013
 

Dernier

ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProduct Anonymous
 
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost SavingRepurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost SavingEdi Saputra
 
Vector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptxVector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptxRemote DBA Services
 
WSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2
 
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 Takeoffsammart93
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWERMadyBayot
 
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 TerraformAndrey Devyatkin
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdfSandro Moreira
 
Introduction to Multilingual Retrieval Augmented Generation (RAG)
Introduction to Multilingual Retrieval Augmented Generation (RAG)Introduction to Multilingual Retrieval Augmented Generation (RAG)
Introduction to Multilingual Retrieval Augmented Generation (RAG)Zilliz
 
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Orbitshub
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...apidays
 
MS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsMS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsNanddeep Nachan
 
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 DiscoveryTrustArc
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century educationjfdjdjcjdnsjd
 
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...Jeffrey Haguewood
 
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...DianaGray10
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodJuan lago vázquez
 
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfRising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfOrbitshub
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 

Dernier (20)

ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost SavingRepurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
Repurposing LNG terminals for Hydrogen Ammonia: Feasibility and Cost Saving
 
Vector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptxVector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptx
 
WSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering Developers
 
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
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
 
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
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf
 
Introduction to Multilingual Retrieval Augmented Generation (RAG)
Introduction to Multilingual Retrieval Augmented Generation (RAG)Introduction to Multilingual Retrieval Augmented Generation (RAG)
Introduction to Multilingual Retrieval Augmented Generation (RAG)
 
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
 
MS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsMS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectors
 
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
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century education
 
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
 
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
 
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...
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
 
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfRising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 

Compressed Sensing - Achuta Kadambi

  • 1. Compressed Sensing: Introduction and Apps Achuta Kadambi Camera Culture, MIT
  • 2. Exploiting Signals • Not all signals are equal! Find a weakness then exploit.
  • 3. Exploiting Signals • Not all signals are equal! Find a weakness then exploit. • Shannon-Nyquist  Bandlimited signals can be sampled/reconstructed
  • 4. Exploiting Signals • Not all signals are equal! Find a weakness then exploit. • Shannon-Nyquist  Bandlimited signals can be sampled/reconstructed • Rank-constrained Optimization  Low Rank signals can be interpolated (Netflix Problem)
  • 5. Exploiting Signals • Not all signals are equal! Find a weakness then exploit. • Shannon-Nyquist  Bandlimited signals can be sampled/reconstructed • Rank-constrained Optimization  Low Rank signals can be interpolated. (Netflix Problem) • Compressed Sensing  Sparse signals can be undersampled and recovered.
  • 6. Outline of this talk. • Compressed Sensing overview. • Very brief explanation on the why and how of Compressed Sensing. • ‘Apps’ that use compressed sensing. • Practical strategies for implementation (e.g. pseudocode, libraries).
  • 7. Motivation: JPEG Compression Our visual system is less sensitive to high (spatial) frequency detail. Can we throw away these frequencies and retain a similar image? This is the intuition behind JPEG. Spatial Frequency E.g.: High  Hair, Blades of Grass, etc. Low  Sky, Skin, etc. Compressed Sensing: If we are going to throw away stuff … why spend time acquiring it?
  • 9. Wired Magazine: “Fill in the Blanks…”
  • 10. 1D Implementation in L1Magic Step 1: The original signal and its Fourier Transform. Original Signal (N = 256) Spectrum
  • 11. Implementation in L1Magic Step 2: The subsampled signal Red Entries (80 samples) are observed. Blue Entries (176 samples) must be recovered. That means we observe only 30% of the original signal.
  • 12. Implementation in L1Magic Step 3: Exact Recovery of the Signal. Original Signal (N = 256) Reconstruction (N = 256)
  • 13. L1Magic for Images Original Image: 1 million pixels Reconstruction: from 100,000 random measurements.
  • 14. Goes back to Fourier
  • 15. Fourier Transform Intuition: Projection, or Inner Product, of Signal with Trigonometric Functions.
  • 16. Sparsity goes back to Fourier (circa 1800) Superposition of Sinusoids Original Time Domain Function Frequency Domain Representation
  • 17. Discrete Fourier Transform Example DFT: Time Signal is a Delta. Spectrum is Broadband.
  • 19. Nyquist-Shannon Sampling Theorem In Shannon’s words: How to Reconstruct? (Interpolation) Compressive Sensing: Can we do better?
  • 20. Inverse Problem Example 1: Sinc Interpolation. Given the Data (a sufficiently sampled signal), how can we obtain the original signal? Example 2: Blurry Photos. Given a Blurry Photo, from a Camera, how can we go back to the original, sharp image? Example 3: Given a discrete time signal, how can we obtain its discrete spectrum? **DFT problem is a Linear Inverse Problem
  • 21. Solving the DFT Problem Done?
  • 22. Solving the DFT Problem via Optimization Done? Loss Function
  • 23. Solving the DFT Problem via Optimization Done? PseudoInverse: Minimize MSE
  • 24. Constraining our Solution via Regularization We can go beyond loss function, e.g., Tikhonov Regularization Additional Term allows for some prior on original signal. For instance if Tikhonov Matrix is a first order difference, then you are biasing x toward smooth solutions. Linked to the Lagrange problem, as well as Maximum A Posteriori from probability, and Weiner filter from Sig proc.
  • 26. Compressed Sensing Structure • Underdetermined system. y=Ax. • • • • Y is m-dimensional sampled vector A is mxn matrix X is n-dimensional original vector. And m << n y A x
  • 27. Simply Solving y=Ax not good enough • This gives you an affine space with many solutions to y=Ax. • So we must constrain our problem to look for the sparse solution to y=Ax.
  • 28. Occams Razor Occam's Razor: among otherwise equal explanations, the simplest is best
  • 29. Occams Razor Occam's Razor: among otherwise equal explanations, the simplest is best CS Occam's Razor: among otherwise equal solutions, the sparsest is best Unfortunately, this optimization is not tractable
  • 31.
  • 33. RIP/Spark/Coherence • The sensing matrix A must be carefully chosen. • For compressed sensing to work, the matrix A must satisfy the Restricted Isometry Property (RIP): • Calculating RIP is NP-hard. We can work with easier quantities than the RIP, such as spark and mutual incoherence.
  • 35. App1: Single-Pixel Camera Design Advantage: A MP camera with just a single-pixel.
  • 36. App2: Single-Pixel THz imaging. Design Advantage: CS allows for single-pixel THz sensors, which are much easier to fabricate than pixel array. In general, you can buy amazing things at single-pixel level, e.g., picosecond detectors, thermal IR sensor, etc. Chan et al. Applied Physics 2008
  • 37. App2: Monitoring Breathing via Smartphone Very similar to OMP, DFT formulation. Basically finding sparse spectral components that characterize the audio signal of breathing. Design Advantage: Compressed sensing allows for low power acquisition and reduced streaming. Oletic, Skrapec, and Bilas MobiHealth 2012
  • 38. App4: Biometrics … Face Recognition Design Advantage: Using compressed sensing to handle the small sample size problem. Before, the number of samples in the database Is less than the degrees of freedom of each sample.
  • 39. App5: Fast MRI Design Advantage: Less samples means less time for an MRI Scan, which means less time a sick or disabled patient lies in the scanner.
  • 41. App6: Compressive Sensing of High Speed Periodic Videos Design Advantage: Exploit sparsity of Periodic Videos to obtain a high speed video without using a high speed camera. Veeraraghavan A, Reddy D, Raskar R. IEEE PAMI
  • 42. App7: Compressive Light Field Photography Marwah, Wetzstein, Bando, Ra skar. ACM SIGGRAPH 2013. Design Advantage: Obtain High-Resolution Light Field photos by placing a coded mask in front of the sensor.
  • 43. App8: Sparsity-Induced Time of Flight Cameras Kadambi et al. ACM SIGGRAPH Asia E.g., Light Sweep Movies from Refael’s talk last week. Goal is to obtain a well-conditioned deconvolution problem. Design Advantage: Deconvolve to obtain bounces of light and construct a light sweep video.
  • 44. App9  YOUR App! Design Advantage: <insert here>
  • 45. Practical Strategies Many libraries are available for C++/Matlab/etc. • Recommended: L1magic (http://users.ece.gatech.edu/~justin/l1magic/) • SPGL1 (http://www.cs.ubc.ca/~mpf/spgl1/) • CVX (http://cvxr.com/cvx/) • On phone, nothing exists yet, but you can use Efficient Java Matrix Library (EJML) to implement solely in linear algebra.
  • 46. L1Magic Pseudocode For 1D signal. x = original_signal; R = randn(m,n); A = orth(R’)’; // read in your original signal. // create a random matrix of dimension mxn // sensing matrix with orthogonal columns. y = A*x; // create subsampled signal y of only m entries. X0 = A’*y; // Initial guess by taking matrix inverse. x_hat = l1eq_pd(x0, A, [], y, 1e-3); // run l1 solver. norm(x_hat – x); // error; should be zero in ideal case
  • 47. Take-home Messages • Opportunity to integrate cutting-edge mathematical techniques into your camera apps. • Compressed Sensing is lightweight in terms of coding. The key is correctly identifying the sparsity in your engineering problem. • Not all signals are equal… find a weakness, e.g., sparsity/rank, and exploit it. • Exploit in Hardware, Exploit in Software…