SlideShare a Scribd company logo
1 of 21
Download to read offline
Network Deconvolution
Choi Dongmin
Yonsei University Severance Hospital CCIDS
• Images of natural scenes contain adjacent pixels that are
statistically highly correlated

• This correlation effect makes network training challenging
because adjacent pixels contain redundant information

• Hubel and Wiesel found that a visual correlation removal process
exists in animal brain
https://www.researchgate.net/figure/Fig-In-the-classic-neuroscience-experiment-Hubel-and-Wiesel-discovered-a-cats-visual_fig1_335707980
Introduction
• Network deconvolution

- a decorrelation method to remove both the pixel-wise and

channel-wise correlation at each layer of the network
Introduction
A correlated signal 

( : kernel / : the corresponding convolution matrix)

Removing the correlation effects via:
b = k * x = Kx
k K
x = K−1
b
• Contributions

- Introduce network deconvolution, a decorrelation method to remove the
both the pixel-wise and channel-wise correlation at each layer of the
network.

- Deconvolution can replace batch normalization with better model training

- Optimal transform if considering optimization

- Deconvolution reduces redundancy in the data, leading to sparse
representations
L2
Introduction
Motivations
LossL2
=
1
2
||y − ̂y||2
=
1
2
||Xw − ̂y||2
( : the inputs, : an unknown weight matrix )X w
wnew = wold − α
1
N
(Xt
Xwold − Xt ̂y)
( : the learning rate )α
A linear regression problem with lossL2
One iteration of gradient descent
Motivations
∂LossL2
∂w
= Xt
(Xw − ̂y) = 0 w = (Xt
X)−1
Xt ̂y
wnew = wold − α
1
N
(Xt
Xwold − Xt ̂y)
An optimal solution (gradient is zero)
Proposition 1. Gradient descent converges to the optimal solution in one iteration if
1
N
Xt
X = I
Eq. 2
Eq. 3
Motivations
Proposition 1. Gradient descent converges to the optimal solution in one iteration if
1
N
Xt
X = I
• : the covariance matrix of the features

- the features should be standardized and uncorrelated with each other

- the more correlated does not hold, the slower the convergence

- the solution for this problem : correcting the gradient by a change of
coordinates so that in the new space we have
1
N
Xt
X = I
1
N
Xt
X = I
The Deconvolution Operation
flattened 2D kernelData matrix created by im2col
Representing standard conv filtering as large matrix multiplicationx * kernel
The Deconvolution Operation
• Given a data matrix , the covariance matrix 

Let and multiply this to (the centered vectors)

Then, the covariance of transformed matrix is 

• That is, the pixel-wise and channel-wise correlation is removed by
multiplying
XN×F Cov =
1
N
(X − μ)T
(X − μ)
D = Cov−1
2 X − μ
(X − μ)·D I
D
DT
(X − μ)T
(X − μ)D = Cov−0.5
·Cov·Cov−0.5
= I
The Deconvolution Operation
• By the associative rule of matrix multiplication,

• Therefore, the deconvolution can be carried out implicitly by changing the
model parameters.

• One training is finished, freeze to be the running average

• This change of parameters makes a network perform faster at testing time

D
y = X·D·w = X·(D·w)
The Deconvolution Operation
Experiments
Linear Regression with loss and Logistic RegressionL2
Experiments
Convolutional Networks on CIFAR-10/100
BN : Batch Normalization / ND : Network Deconvolution / 1,20,100 : trained for epochs
Experiments
Convolutional Networks on ImageNet
Experiments
Generalization To Semantic Segmentation
Source Code
...
https://github.com/yechengxi/deconvolution
Conclusion
• Network deconvolution is likely the correct way of training the
convolutional networks.

• The deconvolution filters resemble the center-surround
structures in animal neurons.
• ICLR2020 Official Blind Review #1 (Rating 8)

- This paper proposes an operation for removing the pixel-wise and channel-wise
correlations of input features

- The approach has a well-sounded neurological inspired motivation

- Achieved a good performance compared with batch normalization

- Providing CPU time is very appreciated



- The computation cost for the im2col in a large kernel (7x7) is insanely large, but
not shown on paper

- The arguments made on the sparse representations is not convincing because
showing only 2 learning curves is not enough
Reviews
• ICLR2020 Official Blind Review #2 (Rating 8)

- Network deconvolution is a generalization of batch normalization that not only
whitens per channel, but also removes correlations between channels and across
spatial locations.



- How about the dependence on batch size? : Reply

- Are results sensitive to the epsilon in algorithm 1? (In computing deconv matrix)

- How does this method interact with regularization methods?
Reviews
Reply : …, our method works best for batch sizes
128/256 on CIFAR10/ImageNet dataset. But it also
works well with relatively small/large batch sizes. ….
When the batch size is tiny, for example 4, we also
need to reduce the learning rate to 0.01 to avoid the
negative effects of noisy samples - this correction is
necessary with batch normalization as well and is
not unique to network deconvolution.
• ICLR2020 Official Blind Review #3 (Rating 6)

- The concept of the paper is pretty simple and straightforward - basically it
removes the correlation present in the input data, specifically in the case of
convolution.



- How about PCA transformation?

: Reply - PCA has a number of issues for whitening

1) Finding the principle axes is slow

2) not well-defined if several axes have the same variance
Reviews
Thank you
Yonsei University Severance Hospital CCIDS

More Related Content

What's hot

PR243: Designing Network Design Spaces
PR243: Designing Network Design SpacesPR243: Designing Network Design Spaces
PR243: Designing Network Design SpacesJinwon Lee
 
PR095: Modularity Matters: Learning Invariant Relational Reasoning Tasks
PR095: Modularity Matters: Learning Invariant Relational Reasoning TasksPR095: Modularity Matters: Learning Invariant Relational Reasoning Tasks
PR095: Modularity Matters: Learning Invariant Relational Reasoning TasksJinwon Lee
 
Convolutional neural network from VGG to DenseNet
Convolutional neural network from VGG to DenseNetConvolutional neural network from VGG to DenseNet
Convolutional neural network from VGG to DenseNetSungminYou
 
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...IOSR Journals
 
PR-284: End-to-End Object Detection with Transformers(DETR)
PR-284: End-to-End Object Detection with Transformers(DETR)PR-284: End-to-End Object Detection with Transformers(DETR)
PR-284: End-to-End Object Detection with Transformers(DETR)Jinwon Lee
 
Exploring Simple Siamese Representation Learning
Exploring Simple Siamese Representation LearningExploring Simple Siamese Representation Learning
Exploring Simple Siamese Representation LearningSungchul Kim
 
PR-317: MLP-Mixer: An all-MLP Architecture for Vision
PR-317: MLP-Mixer: An all-MLP Architecture for VisionPR-317: MLP-Mixer: An all-MLP Architecture for Vision
PR-317: MLP-Mixer: An all-MLP Architecture for VisionJinwon Lee
 
MobileNet - PR044
MobileNet - PR044MobileNet - PR044
MobileNet - PR044Jinwon Lee
 
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...PR-197: One ticket to win them all: generalizing lottery ticket initializatio...
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...Jinwon Lee
 
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...JaeJun Yoo
 
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...Jinwon Lee
 
Super resolution in deep learning era - Jaejun Yoo
Super resolution in deep learning era - Jaejun YooSuper resolution in deep learning era - Jaejun Yoo
Super resolution in deep learning era - Jaejun YooJaeJun Yoo
 
Score based Generative Modeling through Stochastic Differential Equations
Score based Generative Modeling through Stochastic Differential EquationsScore based Generative Modeling through Stochastic Differential Equations
Score based Generative Modeling through Stochastic Differential EquationsSungchul Kim
 
PR-231: A Simple Framework for Contrastive Learning of Visual Representations
PR-231: A Simple Framework for Contrastive Learning of Visual RepresentationsPR-231: A Simple Framework for Contrastive Learning of Visual Representations
PR-231: A Simple Framework for Contrastive Learning of Visual RepresentationsJinwon Lee
 
ShuffleNet - PR054
ShuffleNet - PR054ShuffleNet - PR054
ShuffleNet - PR054Jinwon Lee
 
Review-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learningReview-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learningTrong-An Bui
 
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis taeseon ryu
 
Deep learning lecture - part 1 (basics, CNN)
Deep learning lecture - part 1 (basics, CNN)Deep learning lecture - part 1 (basics, CNN)
Deep learning lecture - part 1 (basics, CNN)SungminYou
 
Parallel convolutional neural network
Parallel  convolutional neural networkParallel  convolutional neural network
Parallel convolutional neural networkAbdullah Khan Zehady
 

What's hot (20)

PR243: Designing Network Design Spaces
PR243: Designing Network Design SpacesPR243: Designing Network Design Spaces
PR243: Designing Network Design Spaces
 
2021 05-04-u2-net
2021 05-04-u2-net2021 05-04-u2-net
2021 05-04-u2-net
 
PR095: Modularity Matters: Learning Invariant Relational Reasoning Tasks
PR095: Modularity Matters: Learning Invariant Relational Reasoning TasksPR095: Modularity Matters: Learning Invariant Relational Reasoning Tasks
PR095: Modularity Matters: Learning Invariant Relational Reasoning Tasks
 
Convolutional neural network from VGG to DenseNet
Convolutional neural network from VGG to DenseNetConvolutional neural network from VGG to DenseNet
Convolutional neural network from VGG to DenseNet
 
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
Using Multi-layered Feed-forward Neural Network (MLFNN) Architecture as Bidir...
 
PR-284: End-to-End Object Detection with Transformers(DETR)
PR-284: End-to-End Object Detection with Transformers(DETR)PR-284: End-to-End Object Detection with Transformers(DETR)
PR-284: End-to-End Object Detection with Transformers(DETR)
 
Exploring Simple Siamese Representation Learning
Exploring Simple Siamese Representation LearningExploring Simple Siamese Representation Learning
Exploring Simple Siamese Representation Learning
 
PR-317: MLP-Mixer: An all-MLP Architecture for Vision
PR-317: MLP-Mixer: An all-MLP Architecture for VisionPR-317: MLP-Mixer: An all-MLP Architecture for Vision
PR-317: MLP-Mixer: An all-MLP Architecture for Vision
 
MobileNet - PR044
MobileNet - PR044MobileNet - PR044
MobileNet - PR044
 
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...PR-197: One ticket to win them all: generalizing lottery ticket initializatio...
PR-197: One ticket to win them all: generalizing lottery ticket initializatio...
 
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...
Rethinking Data Augmentation for Image Super-resolution: A Comprehensive Anal...
 
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...
PR-330: How To Train Your ViT? Data, Augmentation, and Regularization in Visi...
 
Super resolution in deep learning era - Jaejun Yoo
Super resolution in deep learning era - Jaejun YooSuper resolution in deep learning era - Jaejun Yoo
Super resolution in deep learning era - Jaejun Yoo
 
Score based Generative Modeling through Stochastic Differential Equations
Score based Generative Modeling through Stochastic Differential EquationsScore based Generative Modeling through Stochastic Differential Equations
Score based Generative Modeling through Stochastic Differential Equations
 
PR-231: A Simple Framework for Contrastive Learning of Visual Representations
PR-231: A Simple Framework for Contrastive Learning of Visual RepresentationsPR-231: A Simple Framework for Contrastive Learning of Visual Representations
PR-231: A Simple Framework for Contrastive Learning of Visual Representations
 
ShuffleNet - PR054
ShuffleNet - PR054ShuffleNet - PR054
ShuffleNet - PR054
 
Review-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learningReview-image-segmentation-by-deep-learning
Review-image-segmentation-by-deep-learning
 
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
 
Deep learning lecture - part 1 (basics, CNN)
Deep learning lecture - part 1 (basics, CNN)Deep learning lecture - part 1 (basics, CNN)
Deep learning lecture - part 1 (basics, CNN)
 
Parallel convolutional neural network
Parallel  convolutional neural networkParallel  convolutional neural network
Parallel convolutional neural network
 

Similar to Network Deconvolution Removes Correlations for Faster CNN Training

Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...
Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...
Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...Yan Xu
 
Neural Networks in Data Mining - “An Overview”
Neural Networks  in Data Mining -   “An Overview”Neural Networks  in Data Mining -   “An Overview”
Neural Networks in Data Mining - “An Overview”Dr.(Mrs).Gethsiyal Augasta
 
Deep learning for image super resolution
Deep learning for image super resolutionDeep learning for image super resolution
Deep learning for image super resolutionPrudhvi Raj
 
convolutional_neural_networks in deep learning
convolutional_neural_networks in deep learningconvolutional_neural_networks in deep learning
convolutional_neural_networks in deep learningssusere5ddd6
 
Hands on machine learning with scikit-learn and tensor flow by ahmed yousry
Hands on machine learning with scikit-learn and tensor flow by ahmed yousryHands on machine learning with scikit-learn and tensor flow by ahmed yousry
Hands on machine learning with scikit-learn and tensor flow by ahmed yousryAhmed Yousry
 
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptx
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptxEfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptx
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptxssuser2624f71
 
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...ali hassan
 
UNetEliyaLaialy (2).pptx
UNetEliyaLaialy (2).pptxUNetEliyaLaialy (2).pptx
UNetEliyaLaialy (2).pptxNoorUlHaq47
 
Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)Gaurav Mittal
 
ImageNet classification with deep convolutional neural networks(2012)
ImageNet classification with deep convolutional neural networks(2012)ImageNet classification with deep convolutional neural networks(2012)
ImageNet classification with deep convolutional neural networks(2012)WoochulShin10
 
build a Convolutional Neural Network (CNN) using TensorFlow in Python
build a Convolutional Neural Network (CNN) using TensorFlow in Pythonbuild a Convolutional Neural Network (CNN) using TensorFlow in Python
build a Convolutional Neural Network (CNN) using TensorFlow in PythonKv Sagar
 
Cvpr 2018 papers review (efficient computing)
Cvpr 2018 papers review (efficient computing)Cvpr 2018 papers review (efficient computing)
Cvpr 2018 papers review (efficient computing)DonghyunKang12
 
Deep learning for image video processing
Deep learning for image video processingDeep learning for image video processing
Deep learning for image video processingYu Huang
 
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018Universitat Politècnica de Catalunya
 
Autoencoders for image_classification
Autoencoders for image_classificationAutoencoders for image_classification
Autoencoders for image_classificationCenk Bircanoğlu
 

Similar to Network Deconvolution Removes Correlations for Faster CNN Training (20)

Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...
Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...
Deep Learning Approach in Characterizing Salt Body on Seismic Images - by Zhe...
 
ResNet.pptx
ResNet.pptxResNet.pptx
ResNet.pptx
 
ResNet.pptx
ResNet.pptxResNet.pptx
ResNet.pptx
 
Neural Networks in Data Mining - “An Overview”
Neural Networks  in Data Mining -   “An Overview”Neural Networks  in Data Mining -   “An Overview”
Neural Networks in Data Mining - “An Overview”
 
Deep learning for image super resolution
Deep learning for image super resolutionDeep learning for image super resolution
Deep learning for image super resolution
 
convolutional_neural_networks in deep learning
convolutional_neural_networks in deep learningconvolutional_neural_networks in deep learning
convolutional_neural_networks in deep learning
 
Hands on machine learning with scikit-learn and tensor flow by ahmed yousry
Hands on machine learning with scikit-learn and tensor flow by ahmed yousryHands on machine learning with scikit-learn and tensor flow by ahmed yousry
Hands on machine learning with scikit-learn and tensor flow by ahmed yousry
 
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptx
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptxEfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptx
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.pptx
 
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
2017 (albawi-alkabi)image-net classification with deep convolutional neural n...
 
UNetEliyaLaialy (2).pptx
UNetEliyaLaialy (2).pptxUNetEliyaLaialy (2).pptx
UNetEliyaLaialy (2).pptx
 
Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)Convolutional Neural Networks (CNN)
Convolutional Neural Networks (CNN)
 
lec6a.ppt
lec6a.pptlec6a.ppt
lec6a.ppt
 
Deep Learning
Deep LearningDeep Learning
Deep Learning
 
ImageNet classification with deep convolutional neural networks(2012)
ImageNet classification with deep convolutional neural networks(2012)ImageNet classification with deep convolutional neural networks(2012)
ImageNet classification with deep convolutional neural networks(2012)
 
build a Convolutional Neural Network (CNN) using TensorFlow in Python
build a Convolutional Neural Network (CNN) using TensorFlow in Pythonbuild a Convolutional Neural Network (CNN) using TensorFlow in Python
build a Convolutional Neural Network (CNN) using TensorFlow in Python
 
Cvpr 2018 papers review (efficient computing)
Cvpr 2018 papers review (efficient computing)Cvpr 2018 papers review (efficient computing)
Cvpr 2018 papers review (efficient computing)
 
Mnist report
Mnist reportMnist report
Mnist report
 
Deep learning for image video processing
Deep learning for image video processingDeep learning for image video processing
Deep learning for image video processing
 
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018
Convolutional Neural Networks - Veronica Vilaplana - UPC Barcelona 2018
 
Autoencoders for image_classification
Autoencoders for image_classificationAutoencoders for image_classification
Autoencoders for image_classification
 

More from Dongmin Choi

[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...Dongmin Choi
 
Review: Incremental Few-shot Instance Segmentation [CDM]
Review: Incremental Few-shot Instance Segmentation [CDM]Review: Incremental Few-shot Instance Segmentation [CDM]
Review: Incremental Few-shot Instance Segmentation [CDM]Dongmin Choi
 
Review: You Only Look One-level Feature
Review: You Only Look One-level FeatureReview: You Only Look One-level Feature
Review: You Only Look One-level FeatureDongmin Choi
 
Transformer in Computer Vision
Transformer in Computer VisionTransformer in Computer Vision
Transformer in Computer VisionDongmin Choi
 
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...Dongmin Choi
 
YolactEdge Review [cdm]
YolactEdge Review [cdm]YolactEdge Review [cdm]
YolactEdge Review [cdm]Dongmin Choi
 
Review : Inter-slice Context Residual Learning for 3D Medical Image Segmentation
Review : Inter-slice Context Residual Learning for 3D Medical Image SegmentationReview : Inter-slice Context Residual Learning for 3D Medical Image Segmentation
Review : Inter-slice Context Residual Learning for 3D Medical Image SegmentationDongmin Choi
 
Deformable DETR Review [CDM]
Deformable DETR Review [CDM]Deformable DETR Review [CDM]
Deformable DETR Review [CDM]Dongmin Choi
 
ViT (Vision Transformer) Review [CDM]
ViT (Vision Transformer) Review [CDM]ViT (Vision Transformer) Review [CDM]
ViT (Vision Transformer) Review [CDM]Dongmin Choi
 
Review : Prototype Mixture Models for Few-shot Semantic Segmentation
Review : Prototype Mixture Models for Few-shot Semantic SegmentationReview : Prototype Mixture Models for Few-shot Semantic Segmentation
Review : Prototype Mixture Models for Few-shot Semantic SegmentationDongmin Choi
 
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Dongmin Choi
 
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]Dongmin Choi
 
Review : Rethinking Pre-training and Self-training
Review : Rethinking Pre-training and Self-trainingReview : Rethinking Pre-training and Self-training
Review : Rethinking Pre-training and Self-trainingDongmin Choi
 
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...Dongmin Choi
 
Pyradiomics Customization [CDM]
Pyradiomics Customization [CDM]Pyradiomics Customization [CDM]
Pyradiomics Customization [CDM]Dongmin Choi
 
Seeing What a GAN Cannot Generate [cdm]
Seeing What a GAN Cannot Generate [cdm]Seeing What a GAN Cannot Generate [cdm]
Seeing What a GAN Cannot Generate [cdm]Dongmin Choi
 
Neural network pruning with residual connections and limited-data review [cdm]
Neural network pruning with residual connections and limited-data review [cdm]Neural network pruning with residual connections and limited-data review [cdm]
Neural network pruning with residual connections and limited-data review [cdm]Dongmin Choi
 
How much position information do convolutional neural networks encode? review...
How much position information do convolutional neural networks encode? review...How much position information do convolutional neural networks encode? review...
How much position information do convolutional neural networks encode? review...Dongmin Choi
 
Objects as points (CenterNet) review [CDM]
Objects as points (CenterNet) review [CDM]Objects as points (CenterNet) review [CDM]
Objects as points (CenterNet) review [CDM]Dongmin Choi
 
Augmix review [cdm]
Augmix review [cdm]Augmix review [cdm]
Augmix review [cdm]Dongmin Choi
 

More from Dongmin Choi (20)

[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
[Review] BoxInst: High-Performance Instance Segmentation with Box Annotations...
 
Review: Incremental Few-shot Instance Segmentation [CDM]
Review: Incremental Few-shot Instance Segmentation [CDM]Review: Incremental Few-shot Instance Segmentation [CDM]
Review: Incremental Few-shot Instance Segmentation [CDM]
 
Review: You Only Look One-level Feature
Review: You Only Look One-level FeatureReview: You Only Look One-level Feature
Review: You Only Look One-level Feature
 
Transformer in Computer Vision
Transformer in Computer VisionTransformer in Computer Vision
Transformer in Computer Vision
 
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...
Review : Adaptive Consistency Regularization for Semi-Supervised Transfer Lea...
 
YolactEdge Review [cdm]
YolactEdge Review [cdm]YolactEdge Review [cdm]
YolactEdge Review [cdm]
 
Review : Inter-slice Context Residual Learning for 3D Medical Image Segmentation
Review : Inter-slice Context Residual Learning for 3D Medical Image SegmentationReview : Inter-slice Context Residual Learning for 3D Medical Image Segmentation
Review : Inter-slice Context Residual Learning for 3D Medical Image Segmentation
 
Deformable DETR Review [CDM]
Deformable DETR Review [CDM]Deformable DETR Review [CDM]
Deformable DETR Review [CDM]
 
ViT (Vision Transformer) Review [CDM]
ViT (Vision Transformer) Review [CDM]ViT (Vision Transformer) Review [CDM]
ViT (Vision Transformer) Review [CDM]
 
Review : Prototype Mixture Models for Few-shot Semantic Segmentation
Review : Prototype Mixture Models for Few-shot Semantic SegmentationReview : Prototype Mixture Models for Few-shot Semantic Segmentation
Review : Prototype Mixture Models for Few-shot Semantic Segmentation
 
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
Review : PolarMask: Single Shot Instance Segmentation with Polar Representati...
 
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]
Review : Multi-Domain Image Completion for Random Missing Input Data [cdm]
 
Review : Rethinking Pre-training and Self-training
Review : Rethinking Pre-training and Self-trainingReview : Rethinking Pre-training and Self-training
Review : Rethinking Pre-training and Self-training
 
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...
Review : Structure Boundary Preserving Segmentation
for Medical Image with Am...
 
Pyradiomics Customization [CDM]
Pyradiomics Customization [CDM]Pyradiomics Customization [CDM]
Pyradiomics Customization [CDM]
 
Seeing What a GAN Cannot Generate [cdm]
Seeing What a GAN Cannot Generate [cdm]Seeing What a GAN Cannot Generate [cdm]
Seeing What a GAN Cannot Generate [cdm]
 
Neural network pruning with residual connections and limited-data review [cdm]
Neural network pruning with residual connections and limited-data review [cdm]Neural network pruning with residual connections and limited-data review [cdm]
Neural network pruning with residual connections and limited-data review [cdm]
 
How much position information do convolutional neural networks encode? review...
How much position information do convolutional neural networks encode? review...How much position information do convolutional neural networks encode? review...
How much position information do convolutional neural networks encode? review...
 
Objects as points (CenterNet) review [CDM]
Objects as points (CenterNet) review [CDM]Objects as points (CenterNet) review [CDM]
Objects as points (CenterNet) review [CDM]
 
Augmix review [cdm]
Augmix review [cdm]Augmix review [cdm]
Augmix review [cdm]
 

Recently uploaded

How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?XfilesPro
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAndikSusilo4
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 3652toLead Limited
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 
Pigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...shyamraj55
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphNeo4j
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationRidwan Fadjar
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsMark Billinghurst
 
Maximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxMaximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxOnBoard
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Patryk Bandurski
 

Recently uploaded (20)

How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & Application
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
Pigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food ManufacturingPigging Solutions in Pet Food Manufacturing
Pigging Solutions in Pet Food Manufacturing
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 Presentation
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR Systems
 
Maximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxMaximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptx
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
 

Network Deconvolution Removes Correlations for Faster CNN Training

  • 1. Network Deconvolution Choi Dongmin Yonsei University Severance Hospital CCIDS
  • 2. • Images of natural scenes contain adjacent pixels that are statistically highly correlated • This correlation effect makes network training challenging because adjacent pixels contain redundant information • Hubel and Wiesel found that a visual correlation removal process exists in animal brain https://www.researchgate.net/figure/Fig-In-the-classic-neuroscience-experiment-Hubel-and-Wiesel-discovered-a-cats-visual_fig1_335707980 Introduction
  • 3. • Network deconvolution
 - a decorrelation method to remove both the pixel-wise and
 channel-wise correlation at each layer of the network Introduction A correlated signal 
 ( : kernel / : the corresponding convolution matrix)
 Removing the correlation effects via: b = k * x = Kx k K x = K−1 b
  • 4. • Contributions
 - Introduce network deconvolution, a decorrelation method to remove the both the pixel-wise and channel-wise correlation at each layer of the network.
 - Deconvolution can replace batch normalization with better model training
 - Optimal transform if considering optimization
 - Deconvolution reduces redundancy in the data, leading to sparse representations L2 Introduction
  • 5. Motivations LossL2 = 1 2 ||y − ̂y||2 = 1 2 ||Xw − ̂y||2 ( : the inputs, : an unknown weight matrix )X w wnew = wold − α 1 N (Xt Xwold − Xt ̂y) ( : the learning rate )α A linear regression problem with lossL2 One iteration of gradient descent
  • 6. Motivations ∂LossL2 ∂w = Xt (Xw − ̂y) = 0 w = (Xt X)−1 Xt ̂y wnew = wold − α 1 N (Xt Xwold − Xt ̂y) An optimal solution (gradient is zero) Proposition 1. Gradient descent converges to the optimal solution in one iteration if 1 N Xt X = I Eq. 2 Eq. 3
  • 7. Motivations Proposition 1. Gradient descent converges to the optimal solution in one iteration if 1 N Xt X = I • : the covariance matrix of the features
 - the features should be standardized and uncorrelated with each other
 - the more correlated does not hold, the slower the convergence
 - the solution for this problem : correcting the gradient by a change of coordinates so that in the new space we have 1 N Xt X = I 1 N Xt X = I
  • 8. The Deconvolution Operation flattened 2D kernelData matrix created by im2col Representing standard conv filtering as large matrix multiplicationx * kernel
  • 9. The Deconvolution Operation • Given a data matrix , the covariance matrix 
 Let and multiply this to (the centered vectors)
 Then, the covariance of transformed matrix is 
 • That is, the pixel-wise and channel-wise correlation is removed by multiplying XN×F Cov = 1 N (X − μ)T (X − μ) D = Cov−1 2 X − μ (X − μ)·D I D DT (X − μ)T (X − μ)D = Cov−0.5 ·Cov·Cov−0.5 = I
  • 10. The Deconvolution Operation • By the associative rule of matrix multiplication, • Therefore, the deconvolution can be carried out implicitly by changing the model parameters. • One training is finished, freeze to be the running average • This change of parameters makes a network perform faster at testing time
 D y = X·D·w = X·(D·w)
  • 12. Experiments Linear Regression with loss and Logistic RegressionL2
  • 13. Experiments Convolutional Networks on CIFAR-10/100 BN : Batch Normalization / ND : Network Deconvolution / 1,20,100 : trained for epochs
  • 17. Conclusion • Network deconvolution is likely the correct way of training the convolutional networks. • The deconvolution filters resemble the center-surround structures in animal neurons.
  • 18. • ICLR2020 Official Blind Review #1 (Rating 8)
 - This paper proposes an operation for removing the pixel-wise and channel-wise correlations of input features
 - The approach has a well-sounded neurological inspired motivation
 - Achieved a good performance compared with batch normalization
 - Providing CPU time is very appreciated
 
 - The computation cost for the im2col in a large kernel (7x7) is insanely large, but not shown on paper
 - The arguments made on the sparse representations is not convincing because showing only 2 learning curves is not enough Reviews
  • 19. • ICLR2020 Official Blind Review #2 (Rating 8)
 - Network deconvolution is a generalization of batch normalization that not only whitens per channel, but also removes correlations between channels and across spatial locations.
 
 - How about the dependence on batch size? : Reply
 - Are results sensitive to the epsilon in algorithm 1? (In computing deconv matrix)
 - How does this method interact with regularization methods? Reviews Reply : …, our method works best for batch sizes 128/256 on CIFAR10/ImageNet dataset. But it also works well with relatively small/large batch sizes. …. When the batch size is tiny, for example 4, we also need to reduce the learning rate to 0.01 to avoid the negative effects of noisy samples - this correction is necessary with batch normalization as well and is not unique to network deconvolution.
  • 20. • ICLR2020 Official Blind Review #3 (Rating 6)
 - The concept of the paper is pretty simple and straightforward - basically it removes the correlation present in the input data, specifically in the case of convolution.
 
 - How about PCA transformation?
 : Reply - PCA has a number of issues for whitening
 1) Finding the principle axes is slow
 2) not well-defined if several axes have the same variance Reviews
  • 21. Thank you Yonsei University Severance Hospital CCIDS