Closed Mgz-97 closed 4 years ago
I don’t think this is possible because the neural net uses the whole image as an input, so even if you change a couple of pixels the result may be different. It shouldn’t vary too much, but it won’t be the exact same. If you want to compare outputs you can define a template and then compute distance between it and further outputs.
Hello, I modified the program to input a picture, but when a similar picture is input, the coordinates of the key points of the output gesture will be different. How to eliminate this difference? Thank you!