Learn TensorFlow 2.0: Implement Machine Learning and Deep Learning Models with Python by Avinash Manure & Pramod Singh

Learn TensorFlow 2.0: Implement Machine Learning and Deep Learning Models with Python by Avinash Manure & Pramod Singh

Author:Avinash Manure & Pramod Singh [Avinash Manure]
Language: eng
Format: epub
Publisher: Apress
Published: 2019-12-16T16:00:00+00:00


Conclusion

In this chapter, you have seen how easy it is to build neural networks in TensorFlow 2.0 and also how to leverage Keras models, to build custom TensorFlow estimators.

© Pramod Singh, Avinash Manure 2020

P. Singh, A. ManureLearn TensorFlow 2.0https://doi.org/10.1007/978-1-4842-5558-2_4

4. Images with TensorFlow

Pramod Singh1 and Avinash Manure2

(1)Bangalore, Karnataka, India

(2)Bangalore, India

This chapter focuses on how we can leverage TensorFlow 2.0 for computer vision. There has been much breakthrough research and development in the field of computer vision, thanks to deep learning. In this chapter, we will start with a brief overview of image processing and move on to one of the most successful algorithms in computer vision, the convolutional neural networks (CNNs), or ConvNets. We will approach CNNs with an introduction and explain their basic architecture with a simple example. Later in this chapter, we will implement a CNN, using TensorFlow 2.0. We will move on to discuss generative networks, which are networks developed for generating images with machines. We will cover autoencoders and variational autoencoders (VAEs), which are a form of generative network. Next, we will implement VAEs, using TensorFlow 2.0, and generate some new images. In the final section, we will discuss the concept of transfer learning—how it has been leveraged in computer vision and the difference between a typical machine learning process and transfer learning. Finally, we discuss applications and the advantages of transfer learning.



Download



Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.