Deep Learning in the World of Little Ponies

Explore computer vision and image classification in this conference talk focused on building a classifier for "My Little Pony" characters. Learn the fundamentals of deep learning, convolutional neural networks, and transfer learning using Keras and Python. Discover how to find and prepare data, use cognitive services and Custom Vision AI, and implement techniques like feature extraction, neural network layers, and data augmentation. Follow along with demonstrations and practical examples, including a steel model implementation, to gain hands-on experience in applying deep learning concepts to real-world image classification tasks.

Syllabus

Intro
Who is Galiya
Finding data
Cognitive services
Custom Vision AI
Adding images
Training
Testing
How to use it
What is an image
Before Deep Learning
Feature Extraction
Representations
Neurons
Layers of Neural Network
Activation Functions
Karis
Traditional neural network architectures
Neural network layers
Databricks
Image Classification
Visualization
Convolution Layer
Pooling Layer
Feature Map
Fully Connected Layer
Overfitting
Date Augmentation
Dropout
Demo
Transfer Learning
Demonstration
Steel Model