dansontong

> If I'm not mistaken, the code for pointnet2's farthest point sampling is in "SASA\pcdet\ops\pointnet2\pointnet2_batch\pointnet2_utils.py" > > There's a statement in there: from . import pointnet2_batch_cuda as pointnet2 > >...

especially the second case ( 3th,4th,5th result ). kernel_initializer is constant 1. even though kernel was flipped in function conv4d, the result should be the same. anyone can help?

done! **variable 'out_frame' should be turn int by this way:** `out_frame = j - (i - int(l_k/2)) - int((l_i - l_o)/2)` `# flipped, key point: add different signal by differnt...

@glenn-jocher I find this problem is latest version in main branch. when yolov8-obb caculate true-positive(tp) in first line of _process_batch() in ultralytics/models/yolo/obb/val.py, it use 'iou = batch_probiou()', I think this...