Hands-on Deep Learning by Harsh Bhasin

Author:Harsh Bhasin
Language: eng
Format: epub
ISBN: 9798868810350
Publisher: Apress

The next chapter revisits LeNet and compares it with AlexNet. It also discusses why this architecture works wonders with handwritten digits but does not perform well with complex images.

Conclusion

The previous chapters of this book discussed MLP. There are two major problems with these methods:i.

In these networks the number of connections is huge; therefore, the learning requires many inputs and takes time.

ii.

This model does not take into account the spatial correlation.

This chapter introduced the components of Convolutional Neural Networks. It discusses the importance of convolution and presents the implementations of the pooling layer, the convolutional layer, etc. Also, the convolution operation explained is slightly different from the mathematical convolution.

The reader should be able to implement the layers from scratch using NumPy after reading this chapter. Also, the reader is expected to appreciate the importance of multiple kernels in CNN. However, one need not implement everything from scratch; Keras provides the implementations of all the layers. The next chapter introduces Keras and explains the implementation of layers using Keras. The chapter also introduces some of the most important CNNs and their implementations. It will empower you with the most powerful weapons to fight the problems of image analysis. To conclude, we started with neurocognition. The following image (Figure 6-9) of neurocognition has been generated by AI (https://​gencraft.​com/​generate) and uses CNN.

Figure 6-9Image of neurocognition generated by https://​gencraft.​com/​generate

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.