How to Effectively Read Research Papers Like RCNN, Fast RCNN, etc.?

Hey everyone,

I'm starting to read research papers on object detection, beginning with RCNN, Fast RCNN, and others. Since these papers are packed with technical details, I want to make sure I'm reading them the right way and extracting the most useful insights.

For those experienced in reading ML/AI research papers, what’s the best approach? Specifically:

What key things should I focus on while reading these papers!?

Any strategies to understand the important contributions of a paper?

I’d appreciate any tips or resources that helped you when you started! Thanks in advance.