Deep Learning for Computer Vision
Lecture 3
Language has transformed the way that we interact with deep learning
Slide 4
Slide 5
What is Deep Learning?
Software Labs: TensorFlow and PyTorch
Why Deep Learning and Why Now?
Why Deep Learning?
Why Now?
The Perceptron
The Perceptron: Forward Propagation
The Perceptron: Forward Propagation
The Perceptron: Forward Propagation
The Perceptron: Forward Propagation
Common Activation Functions
Importance of Activation Functions
Importance of Activation Functions
Importance of Activation Functions
The Perceptron: Example
The Perceptron: Example
The Perceptron: Example
The Perceptron: Example
Building Neural Networks with Perceptrons
The Perceptron: Simplified
The Perceptron: Simplified
Multi Output Perceptron
Dense layer from scratch
Multi Output Perceptron
Single Layer Neural Network
Single Layer Neural Network
Multi Output Perceptron
Deep Neural Network
Deep Neural Network
Applying Neural Networks
Example Problem
Example Problem: Will I pass this class?
Example Problem: Will I pass this class?
Example Problem: Will I pass this class?
Example Problem: Will I pass this class?
Quantifying Loss
Empirical Loss
Binary Cross Entropy Loss
Mean Squared Error Loss
Training Neural Networks
Loss Optimization
Loss Optimization
Loss Optimization
Loss Optimization
Loss Optimization
Loss Optimization
Gradient Descent
Gradient Descent
Gradient Descent
Gradient Descent
Computing Gradients: Backpropagation
Computing Gradients: Backpropagation
Computing Gradients: Backpropagation
Computing Gradients: Backpropagation
Computing Gradients: Backpropagation
Computing Gradients: Backpropagation
Neural Networks in Practice: Optimization
Training Neural Networks is Difficult
Loss Functions Can Be Difficult to Optimize
Loss Functions Can Be Difficult to Optimize
Setting the Learning Rate
Setting the Learning Rate
Setting the Learning Rate
Howtodealwiththis?
How to deal with this?
Adaptive Learning Rates
Gradient Descent Algorithms
Putting it all together
Neural Networks in Practice: Mini-batches
Gradient Descent
Gradient Descent
Stochastic Gradient Descent
Stochastic Gradient Descent
Stochastic Gradient Descent
Stochastic Gradient Descent
Mini-batches while training
Mini-batches while training
Neural Networks in Practice: Overfitting
The Problem of Overfitting
Regularization
Regularization
Regularization I: Dropout
Regularization I: Dropout
Regularization I: Dropout
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Regularization 2: Early Stopping
Core Foundation Review
