I find your article very interesting. Especially, the point that you have mentioned about incorporating modifications to existing models. I have been trying to work on something similar but could not really find the right person to discuss with.
When you had mentioned, “incorporating the modifications in prevailing models such as Yolo for real time object detection” and this is exactly what I’m trying to do. Could you please explain/ elaborate this idea. I am looking for making modifications to YOLO with respect to the dynamic routing algorithm from CapsNet but clueless as to how to go about it.
Your inputs are highly valuable for my project completion.