Hi, I go over your code in details recently, but in function add_virtual_mask, i find it compute the distance between a virtual point and all real point then select the minimum, ignore if the real point is in the same camera_id and same mask with the virtual point. I wonder that's a little unreasonable? Because it may use a real point's depth which is in other mask or even in other camera_id. Looking forward to your reply~
Hi, I go over your code in details recently, but in function add_virtual_mask, i find it compute the distance between a virtual point and all real point then select the minimum, ignore if the real point is in the same camera_id and same mask with the virtual point. I wonder that's a little unreasonable? Because it may use a real point's depth which is in other mask or even in other camera_id. Looking forward to your reply~