The most approachable introduction to DL is a combination of deeplearning.ai and fast.ai. Use fast.ai as a practical bottom-up course and deeplearning.ai to build on the theoretical aspects taught in the fast.ai course.
For people who are looking for a more mathematical introduction, NYU's Deep Learning course is the perfect choice. It teaches theory and offers coding practice in a really intuitive but also rigorous way.
The most comprehensive book on the foundational topics in deep learning. For companion videos of most of the chapters, check out this YouTube playlist.
A set of introductory lectures on Deep Learning by Andrew Ng with a focus on practical issues such as reading research papers, career advice and ML in healthcare.
Also a great course, but not as comprehensive as the ones in the course list. These lectures can be a great first exposure to the topics and also have great guest lectures.
An older course on deep learning from 2013, which covers topics that are not usually covered in more recent courses (e.g. conditional random fields, restricted Boltzman machines and sparse coding).