Mastering Neural Network Computer Vision with TensorFlow and Keras provides a comprehensive guide to using TensorFlow and Keras for computer vision applications. The book enables readers to develop and exercise the skills needed to use sophisticated pre-trained computer vision models, build simple and more advanced neural network models, and optimize their performance.
The different chapters of the book cover a comprehensive range of topics in computer vision and deep learning. The first chapter provides a theoretical introduction to computer vision and deep learning, and the second one provides an overview of TensorFlow and its capabilities. The subsequent chapters cover specific applications of neural networks in computer vision, such as image classification, image segmentation, and object detection, and how to tap into the power of transfer learning and pre-trained models to address those use cases. Finally, the remaining chapters cover how to design your own neural network, gather a proper dataset and train your model efficiently. They also cover image generation and ethical considerations around computer vision.
By the end of this book, readers will have a strong understanding of the principles of deep learning and computer vision, as well as the skills needed to build advanced neural network models using TensorFlow.
KEY FEATURES
* Master computer vision fundamentals through hands-on implementation with Tensorflow, from basics to advanced applications.
* Learn real-world techniques for preparing data, training models, and deploying computer vision solutions at scale.
* Explore state-of-the-art techniques, including transfer learning, generative models, and advanced vision tasks through practical projects.
WHAT YOU WILL LEARN
* Understand essential deep learning concepts and architectures specifically designed for modern computer vision applications.
* Build practical expertise with Tensorflow and Keras while implementing pre-trained models for vision tasks.
* Learn to fine-tune existing models and design new architectures for specific vision challenges.
* Master techniques to improve model efficiency, training speed, and overall performance in real applications.
* They will know how diffusion-based models work and how to use some of the most popular ones, like DALL-E or Stable Diffusion.
WHO THIS BOOK IS FOR
This book is for current or aspiring deep tech professionals, students, and anyone who wishes to understand the rewarding field of computer vision. More specifically, it will also have a great impact on computer vision engineers, robotics, image processing, and video processing engineers who are willing to learn how to use neural networks to boost their performance and results.
Jean Anoma is an accomplished lecturer and practitioner in the field of neural networks with more than 6 years of practical experience solving complex problems in computer vision and natural language processing. After completing his master's thesis on fast bib recognition with deep learning at the Femto ST research laboratory in France, he held lead positions on high stakes artificial intelligence projects in large multinational companies, notably at Faurecia (now Forvia) on use cases related to visual inspection and then at Amadeus on use cases related to natural language processing. He has developed his pedagogy through numerous interventions as a professional trainer on neural network and computer vision for large training organizations (such as Orsys in France), directly for institutional clients such as Axa Assurances and the African Development Bank Group, and in leading engineering schools like EPITA in France. Jean holds a Master’s degree in computer science from Franche Comte University and a master's degree in mathematics from Sorbonne University. Besides, he holds a master’s degree in management from HEC Paris as well. He won several academic contests during his high school education, having been nominated best twelfth-grade student of his country.
