0% found this document useful (0 votes)
8 views30 pages

ImageNet CNN Classification Insights

The document discusses the architecture and training of a deep convolutional neural network (CNN) for image classification using the ImageNet dataset, highlighting the importance of depth, data augmentation, and techniques to reduce overfitting. It details the use of ReLU nonlinearity, dropout, and local response normalization, along with the training configuration and results achieved in various competitions. The findings emphasize the effectiveness of large, deep networks in supervised learning without unsupervised pre-training.

Uploaded by

minhahsaleem1
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
8 views30 pages

ImageNet CNN Classification Insights

The document discusses the architecture and training of a deep convolutional neural network (CNN) for image classification using the ImageNet dataset, highlighting the importance of depth, data augmentation, and techniques to reduce overfitting. It details the use of ReLU nonlinearity, dropout, and local response normalization, along with the training configuration and results achieved in various competitions. The findings emphasize the effectiveness of large, deep networks in supervised learning without unsupervised pre-training.

Uploaded by

minhahsaleem1
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

ImageNet Classification with Deep Convolutional Neural Networks

Alex Krizhevsky, Ilya Sutskever, Geoffrey


Hinton

무하메드 (2023120216), 티머만크리스토프 (2023126767), 민하살림 (2023120344)


01 Introduction
02 The Architecture
03 Reducing Overfitting
04 Details of learning
05 Results
06 Conclusion and Discussion
Introduction

 Evolution of Realistic Settings: Object classification faces


increased variability in realistic settings.
 Dataset Scale: Larger training sets are imperative, with datasets
like ImageNet boasting 15 million
Q&A labeled high-resolution images
across 22,000 categories.
 CNN Flexibility: CNNs offer scalability through varying depth and
breadth, accommodating diverse datasets effectively.
 GPU Optimization: Optimized GPU implementation of 2D
convolution allows for training large CNNs on sizable datasets
without severe overfitting.
 Training Infrastructure: Two GTX 580 3GB GPUs, coupled with
optimized convolution, reduce training time to 5-6 days.
Introduction

 Brief Overview of Image Classification

Q&A
The Architecture

 ReLU Nonlinearity

- Neurons often modeled with saturating nonlinearities like tanh(x)


or sigmoid f(x) = tanh(x) or f(x) = (1 + e−x)−1.
Q&A
- In gradient descent, these saturating nonlinearities are much
slower to train than the used ReLU function

- The iterations needed to reach 25% training error on CIFAR-10,


highlighting the efficiency of ReLUs.

- Enables experimentation with larger neural networks on datasets


like CIFAR-10.
The Architecture

 Training on Multiple GPUs


 GPU Memory Limitation: Single GTX 580 GPU has 3GB memory, restricting network
size.

 Expanding Network Size: 1.2 million training examples prove sufficient for networks
Q&A
too large for a single GPU.

 Two-GPU Configuration: Utilizing two GPUs for parallelization to overcome memory


constraints.

 Direct GPU Communication: GPUs can read from and write to each other's memory
directly, optimizing parallelization.

 Communication Strategy: Kernels distributed across GPUs with selective


communication, enhancing efficiency.
The Architecture

 Local Response Normalization


 Local Response Normalization (LRN) is a
technique used in the context of neural
networks
 It enhances the generalization and performance of the Q&A
network by mitigating values becoming extremely large
or small
 Each neuron is normalized based on its adjacent
neurons in the layer
 A sum of squares is calculated using a gaussian-like
weighing function. The original neuron value is divided
by this sum, which normalizes the neuron’s response
 Additionally, scaling and bias parameters are added to
enable optimization during training
The Architecture

 Local Response Normalization

Q&A
The Architecture

 Overlapping Pooling

Q&A
The Architecture

 Overall Architecture

Q&A
The Architecture

 Comprehensive Architecture of the


CNN

 Layer Composition: Eight layers with


weights, comprising five convolutional
and three fully-connected layers. Q&A
 Softmax Output: The final fully-
connected layer feeds into a 1000-way
softmax for class label distribution.
 Objective Function: Maximizes the
multinomial logistic regression
objective.
 Training Objective: Equivalent to
maximizing the log-probability of
correct labels across training cases.
The Architecture

 Details of Convolutional Layers

 First Convolutional Layer: 96 kernels of


size 11x11x3 with a stride of 4 pixels.
 Second Convolutional Layer: 256
kernels of size 5x5x48 applied to the Q&A
(response-normalized and pooled)
output of the first layer.
 Third-Fifth Convolutional Layers:
Interconnected without pooling or
normalization, with varying kernel
sizes and connectivity.
 The fully-connected layers have 4096 neurons
each
Reducing Overfitting

 What is
Overfitting?
-Overfitting occurs when a machine
learning model learns the training data
Söhne
too well.
- It often happens with complex models
or when the training dataset is small.
- Including too many features in the
model, especially irrelevant or noisy
features, can lead to overfitting.
Reducing Overfitting

 Data Augmentation Strategies

- Overfitting Challenge: Mitigating overfitting on image data through


artificial dataset enlargement.
Q&A
- Common Method: Label-preserving transformations applied for dataset
augmentation.
- Two Forms of Augmentation: Computationally efficient, generated in
Python code on the CPU during GPU training.
- Real-time Transformation: Augmented images generated in real-time
during GPU training, minimizing storage requirements.
Reducing Overfitting

 Image translations and horizontal reflections

- Image Translations and Reflections: Extracting random 224x224 patches and their
horizontal reflections from 256x256 images.
Q&A
- Training Set Expansion: Increases the training set size by a factor of 2048.
- Overcoming Overfitting: Essential for preventing substantial overfitting, enabling the use
of larger networks.
- Test-time Prediction: Network predictions averaged over ten patches (five and their
horizontal reflections).
Reducing Overfitting

 Intensity Alteration with PCA

- RGB Channel Intensity Alteration: Performing PCA on ImageNet training


set RGB pixel values.
Q&A
- Proportional Magnitudes: Adding multiples of principal components
based on eigenvalues and random variables.
- Illumination Variations: Capturing natural image property of object
identity invariance to intensity changes.
- Error Rate Reduction: This scheme reduces the top-1 error rate by over
1%.
Reducing Overfitting

 What is Dropout?

- Dropout acts as a form of regularization,


preventing overfitting and improving the model's
ability to generalize to new data.
Q&A
- During each training iteration, a random subset
of neurons is dropped out or set to zero. This
means that different sets of neurons are dropped
out in each iteration.

- During each training iteration, a dropout


probability (typically between 0.2 and 0.5) is
chosen.
Reducing Overfitting

 Dropout
Dropout layer

- This Dropout technique randomly sets the output of


each hidden neuron to zero, with a probability of 0.5​
Q&A
- Every time an input is presented, the neural network
samples a different architecture, but they all share the
same weights
- Applied to the first two fully-connected layers of the
neural network.
Details of learning

 Training Configuration

- Optimization Method: Stochastic Gradient


Descent (SGD) utilized for training.

- Batch Size: SGD performed with a batch Q&A


size of 128 examples.

- Momentum: Momentum set to 0.9 for


smoother convergence.
96 convolutional kernels of size 11×11×3 learned by the first convolutional
- Weight Decay: Weight decay of 0.0005 layer on the 224×224×3 input images. The top 48 kernels were learned on
applied, crucial for learning and reducing GPU 1 while the bottom 48 kernels were learned on GPU 2.
training error.
Details of learning

 Weight Update Rule

Q&A

96 convolutional kernels of size 11×11×3 learned by the first


convolutional layer on the 224×224×3 input images. The top 48 kernels
were learned on GPU 1 while the bottom 48 kernels were learned on
GPU 2
Details of learning

 Weight Initialization Strategy

- Gaussian Distribution: Weights in each


layer initialized from a zero-mean Gaussian
distribution with standard deviation $0.01$.
Q&A
- Biases Initialization: Biases in specific
layers initialized differently to accelerate
early learning stages.

96 convolutional kernels of size 11×11×3 learned by the first convolutional


layer on the 224×224×3 input images. The top 48 kernels were learned on
GPU 1 while the bottom 48 kernels were learned on GPU 2
Details of learning

 Learning Rate Adjustment

- Equal Learning Rate: Equal learning rate


applied to all layers, manually adjusted
during training.
Q&A
- Manual Adjustment: Learning rate
adjusted manually based on validation
error rate.

- Learning Rate Reduction: When validation


96 convolutional kernels of size 11×11×3 learned by the first convolutional
error rate plateaus, learning rate divided by layer on the 224×224×3 input images. The top 48 kernels were learned on
$10$. GPU 1 while the bottom 48 kernels were learned on GPU 2

- Initialization: Learning rate initialized at


$0.01$ and reduced three times during
training.
Details of learning

 Training Duration and


Hardware Used
- Training Duration: Network trained for
approximately 90 cycles on 1.2 million
images.
Q&A
- Hardware Configuration: Training
performed on two NVIDIA GTX 580 3GB
GPUs.

- Time Taken: Training process took five to


96 convolutional kernels of size 11×11×3 learned by the first convolutional
six days. layer on the 224×224×3 input images. The top 48 kernels were learned on
GPU 1 while the bottom 48 kernels were learned on GPU 2
Results
 ILSVRC-2010 results

Q&A
Results
 ILSVRC-2012 results

Q&A
Results

 Fall 2009 ImageNet

- 10,184 categories and 8.9 million images

- Half the images are used for training, half for testing
Q&A
- Splits differ from other authors, which does not affect results

- Top-1 and top-5 error rates are 67.4% and 40.9%

- The net above is used, but with an additional sixth convolutional layer over the last pooling layer

- The best published results were 78.1% and 60.9%


Results

 Qualitative Evaluations

Q&A
Conclusion and Discussion

Supervised Learning Success: The paper demonstrates that a large, deep convolutional neural network can
achieve groundbreaking results on a highly challenging dataset through purely supervised learning.

Importance of Depth: The significance of depth in the network's performance is emphasized. Removing even a
single convolutional layer results in a notable loss (approximately
Q&A 2%) in the top-1 performance of the network.
This highlights the importance of the network's depth in achieving the reported results.

Impact of Middle Layers: Specifically, removing any of the middle layers of the network leads to a reduction in
top-1 performance, underscoring the importance of these layers in the learning process.

Absence of Unsupervised Pre-training: Despite acknowledging the potential benefits of unsupervised pre-
training, the authors simplify their experiments by not using it. They express the expectation that unsupervised
pre-training could be particularly helpful with increased computational power, allowing for a larger network
without a proportional increase in labeled data.
Conclusion and Discussion

- In conclusion, this paper presents a pioneering Convolutional


Neural Network (CNN) architecture, showcasing its remarkable
efficacy in object classification tasks. The integration of novel
elements such as rectified linear units
Q&A(ReLUs), dropout for
regularization, and strategic data augmentation contributes to the
CNN's outstanding performance
Thank you for
Q&A

attention

You might also like