index = cvcuda.nms(src, score)
How to use index as a slicing mask to filter src, such as the slicing operation tmp=src [index] similar to numpy/torch
But currently, I have only found that converting the index to torch format tensor is necessary for subsequent operations in the CV-CUDA/samples/object_detection/python/ipelines.py file. How can I implement this box filtering operation solely through cvCUDA
index = cvcuda.nms(src, score)
How to use index as a slicing mask to filter src, such as the slicing operation tmp=src [index] similar to numpy/torch
But currently, I have only found that converting the index to torch format tensor is necessary for subsequent operations in the CV-CUDA/samples/object_detection/python/ipelines.py file. How can I implement this box filtering operation solely through cvCUDA