Overview
Dive into the world of deep learning and convolutional neural networks in this comprehensive 57-minute conference talk from Devoxx. Explore the latest advancements in machine learning, including image recognition and neural network architectures, without needing a PhD. Learn how to apply these cutting-edge technologies to solve previously "impossible" problems in computer science. Gain valuable insights, engineering best practices, and practical tips for implementing deep learning techniques in your own projects. Discover the inner workings of neural networks, TensorFlow code, and hyperparameter tuning. Examine popular architectures like Inception, SqueezeNet, and Darknet, and understand their real-world applications. Witness a live demo showcasing the power of modern convolutional neural nets. Whether you're a seasoned developer or new to machine learning, this talk provides a solid foundation for leveraging deep learning in your work.
Syllabus
Introduction
Why I dont read papers
The reality of deep learning
What is a neural network
Tensorflow code
Rayleigh function
Convolutional Neural Networks
Hyperparameter tuning
Estimator
Inception paper
Global average pooling
Squeezenet
Darknet
Results
How does it work
The results
Demo
Conclusion
Taught by
Devoxx