search

zztianzz / PF-Net-Point-Fractal-Network

CVPR2020 PF-Net: Point Fractal Network for 3D Point Cloud Completion
MIT License
321 stars 64 forks source link

ModuleNotFoundError: No module named 'model_recon_3layers_robustness' #10

Open lrrll opened 4 years ago

lrrll commented 4 years ago

when i run "show_FPNet.py" i don't know why it is wrong, please tell me how to solve it, thanks.

zztianzz commented 4 years ago

Use the latest file 'show_recon.py' to avoid this error.

lrrll commented 4 years ago

thank you, but when i run 'show_recon.py' it came the problem following: Namespace(batchSize=1, beta1=0.9, crop_point_num=512, cropmethod='random_center', cuda=False, dataroot='dataset/train', drop=0.2, each_scales_size=1, learning_rate=0.0002, manualSeed=None, nc=3, netD='', netG='Checkpoint/point_netG.pth', ngpu=2, niter=300, num_scales=3, pnum=2048, point_scales_list=[2048, 1024, 512], weight_decay=0.001, workers=2, wtl2=0.9) Traceback (most recent call last): File "show_recon.py", line 64, in test_dset = shapenet_part_loader.PartDataset( root='./dataset/shapenetcore_part/shapenetcore_partanno_segmentation_benchmark_v0/',classification=True, class_choice='Airplane', npoints=opt.pnum, split='test') File "/home/ouc/lrr/PF-Net-Point-Fractal-Network-master/shapenet_part_loader.py", line 24, in init with open(self.catfile, 'r') as f: FileNotFoundError: [Errno 2] No such file or directory: './dataset/shapenetcore_part/shapenetcore_partanno_segmentation_benchmark_v0/synsetoffset2category.txt' but that .txt is really exist

lrrll commented 4 years ago

'shape_part_loader.py' line 13:dataset_path=os.path.abspath(os.path.join(BASE_DIR, '../dataset/shapenet_part/shapenetcore_partanno_segmentation_benchmark_v0/')) 'show_recon.py' line 64:test_dset = shapenet_part_loader.PartDataset( root='./dataset/shapenetcore_part/shapenetcore_partanno_segmentation_benchmark_v0/',classification=True, class_choice='Airplane', npoints=opt.pnum, split='test')

so i don't know where the problem is

zztianzz commented 4 years ago

I didn't meet this problem before. Can you train the network? Maybe you can copy the path from the train file.