After a recap of the concept of convolution and the classification of object recognition strategies, in this chapter, we have been implementing and training CNNs for object recognition using different languages (Python and Scala) and different open source frameworks (Keras and TensorFlow in the first case, DL4J, ND4J, and Apache Spark in the second) in a hands-on manner.
In the next chapter, we are going to implement a full image classification web application which, behind the scenes, uses a combination of Keras, TensorFlow, DL4J, ND4J, and Spark.
