All About Neural Networks

The world of machine learning is vast, but one area that is experiencing a lot of growth is neural networks. Neural networks learn to accomplish their tasks by reading training data and adjusting their neuron weights to improve their chance of choosing the correct answer.

There are a lot of different kinds of neural networks, and each kind is specialized to work with different kinds of problems. The jetson-inference repository gives you access to a variety of different networks that have already been trained and optimized for different tasks.

Take a look at the networks available in the model-downloader here.

You might have noticed that there are a lot of different networks for different tasks. You have already downloaded some of the basic models when you built the project from source, but you can explore what else is available, and think about what might be useful for you to use in your project, as you learn about the different models during today's lesson.