search

charlesq34 / pointnet2

PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
Other
3.06k stars 894 forks source link

Shapenet Part Segmentation Test Accuracy #75

Open zaiweizhang opened 5 years ago

zaiweizhang commented 5 years ago

Hi,

I have tried to reproduce your result on shapenet part segmentation task but I can only get around 84.7%. Would you mind providing a model, which can reproduce the results you mentioned in the paper?

Thank you!

MoranChen commented 5 years ago

Hello, I tried evaluate.py in part_seg and get eval mean mIoU 0.837075, but I cannot run test.py. It said 'No module named pointnet2_part_seg' . Can you help me solve the problem?

Thank you!

leejiajun commented 5 years ago

@MoranChen same as you

shengrongjin commented 5 years ago

@MoranChen @leejiajun I had the same problem with you ! Only get eval mean mIoU (all shapes): 0.847109 by evaluate.py under the part_seg . However, can't run test.py . Have you solved this problem . If have , can you tell me how to solve it ?

I would be grateful.

QenPion commented 5 years ago

You have to add the rigth subfolder in test.py after the imports:

BASE_DIR2 = '/models' sys.path.append(BASE_DIR2) sys.path.append(os.path.join(BASE_DIR2, 'models')) sys.path.append(os.path.join(BASE_DIR2, 'utils'))

sheshap commented 4 years ago

I could reproduce only

Trial1: eval mean mIoU: 0.816804 eval mean mIoU (all shapes): 0.847668

Trial2: eval mean mIoU: 0.814602 eval mean mIoU (all shapes): 0.848937

Trial3: eval mean mIoU: 0.816642 eval mean mIoU (all shapes): 0.849952