Hi,
I liked your "Data Augmentation for Object Detection via Progressive and Selective Instance-Switching" paper, so I decided to implement some of your methods. I have some questions about selective re-sampling strategy.
Here, "image I" refers to either IA or IB image that is part of the quadruple {IA, IB, iA, iB} of the specific class c.
I understand that i have to iterate over each of quadruple and if any of these images does not satisfy the above equation, so drop that image and as consequence the whole quadruple.
Could you tell me please if it is correct?
Hi, I liked your "Data Augmentation for Object Detection via Progressive and Selective Instance-Switching" paper, so I decided to implement some of your methods. I have some questions about selective re-sampling strategy.
Here, "image I" refers to either IA or IB image that is part of the quadruple {IA, IB, iA, iB} of the specific class c. I understand that i have to iterate over each of quadruple and if any of these images does not satisfy the above equation, so drop that image and as consequence the whole quadruple. Could you tell me please if it is correct?