Hi, I'm trying to make a dog and cat real-time detection application. I am using OpenCV 4.20.
I trained my network using these config files.The network was trained on 227x227 input I managed to get accuracy = 0.93, loss = 0.13 at 10,000 iterations. Then I tried to recognize objects from the stream by simply repeating the following code given below in the answer. The problem is that the outline drawing is in a completely different place. I made sure that blobFromImage is set to Size (227; 227).
I would be grateful for any help