Why there is the need to train CNN and LSTM separately before combining them?

Looking for example about the possibility of using CNN+LSTM classification i went through this example:

https://it.mathworks.com/help/deeplearning/ug/classify-videos-using-deep-learning.html

And i was asking myself why LSTM is trained separately and then combined with convolutional layers pretrained from googleNET?

It is not possible to build the entire architecture from scratch and trained it as whole ?

Answer 1

[Srivardhan Gadila]

Fine-tuning a network built from pretrained network is usually much faster and easier than training a network with randomly initialized weights from scratch.Hence you can take a pretrained image classification network that has already learned to extract powerful and informative features from natural images and use it as a starting point to learn a new task.

Refer to Get Started with Transfer Learning, Pretrained Deep Neural Networks.