sanjana7395 / face_segmentation Goto Github PK

Hi there,

Much appreciated for your work. However, I tried both using your pretrained model and training one from scratch. The testing result of either one does not look right. Below is the predicted result:

Any idea on why this happens? Thanks in advance :)

[INFO] Generating masks...
[INFO] Splitting data into train, validation and test...
Traceback (most recent call last):
File "/home/bl1027/face_segmentation/preprocessing/load_data.py", line 126, in
main()
File "/home/bl1027/face_segmentation/preprocessing/load_data.py", line 114, in main
train, valid, test = split_data(train, valid, test)
File "/home/bl1027/face_segmentation/preprocessing/load_data.py", line 70, in split_data
group_id = splitting.iloc[celebA]['group']
File "/home/bl1027/miniconda3/envs/frp/lib/python3.6/site-packages/pandas/core/indexing.py", line 879, in getitem
return self._getitem_axis(maybe_callable, axis=axis)
File "/home/bl1027/miniconda3/envs/frp/lib/python3.6/site-packages/pandas/core/indexing.py", line 1496, in _getitem_axis
self._validate_integer(key, axis)
File "/home/bl1027/miniconda3/envs/frp/lib/python3.6/site-packages/pandas/core/indexing.py", line 1437, in _validate_integer
raise IndexError("single positional indexer is out-of-bounds")
IndexError: single positional indexer is out-of-bounds

Do you have any pretrained networks to share with us?

I run script "python load_data.py" but file list_eval_partition.txt doesn't exist. Could you share this file?

some file with name list_eval_partition is mentioned in load_data.py which does't exist in the repo. Also, are Celeb_A and Celeb_HQ are different or same, if different then where is celebA present?

Hi dear
Thank you for your project.
when I run train.py script, it start for training and prints these lines then go to pause and don't print anything.

`
1/Unknown - 13s 13s/step - loss: 3.1658 - acc: 0.0616 - mIoU: 0.01282020-11-04 11:56:42.866834: I tensorflow/core/platform/default/device_tracer.cc:641] Collecting 1278 kernel records, 234 memcpy records.

  2/Unknown - 13s 7s/step - loss: 3.1030 - acc: 0.0730 - mIoU: 0.0139 
  3/Unknown - 14s 5s/step - loss: 3.0520 - acc: 0.0800 - mIoU: 0.0149
  4/Unknown - 14s 4s/step - loss: 3.0151 - acc: 0.0868 - mIoU: 0.0158
  5/Unknown - 15s 3s/step - loss: 2.9743 - acc: 0.0991 - mIoU: 0.0177
  6/Unknown - 15s 2s/step - loss: 2.9408 - acc: 0.1135 - mIoU: 0.0196
  7/Unknown - 15s 2s/step - loss: 2.9064 - acc: 0.1288 - mIoU: 0.0217
  8/Unknown - 16s 2s/step - loss: 2.8769 - acc: 0.1451 - mIoU: 0.0240
  9/Unknown - 16s 2s/step - loss: 2.8478 - acc: 0.1617 - mIoU: 0.0262
 10/Unknown - 16s 2s/step - loss: 2.8215 - acc: 0.1782 - mIoU: 0.0285
 11/Unknown - 17s 2s/step - loss: 2.7944 - acc: 0.1972 - mIoU: 0.0311
 12/Unknown - 17s 1s/step - loss: 2.7659 - acc: 0.2168 - mIoU: 0.0337
 13/Unknown - 18s 1s/step - loss: 2.7429 - acc: 0.2321 - mIoU: 0.0358
 14/Unknown - 18s 1s/step - loss: 2.7188 - acc: 0.2492 - mIoU: 0.0382
 15/Unknown - 18s 1s/step - loss: 2.6954 - acc: 0.2673 - mIoU: 0.0409
 16/Unknown - 19s 1s/step - loss: 2.6758 - acc: 0.2847 - mIoU: 0.0432
 17/Unknown - 19s 1s/step - loss: 2.6534 - acc: 0.3008 - mIoU: 0.0455
 18/Unknown - 19s 1s/step - loss: 2.6323 - acc: 0.3146 - mIoU: 0.0474
 19/Unknown - 20s 1s/step - loss: 2.6122 - acc: 0.3281 - mIoU: 0.0494
 20/Unknown - 20s 1s/step - loss: 2.5933 - acc: 0.3395 - mIoU: 0.0512
 21/Unknown - 21s 978ms/step - loss: 2.5727 - acc: 0.3521 - mIoU: 0.0529
 22/Unknown - 21s 951ms/step - loss: 2.5521 - acc: 0.3646 - mIoU: 0.0546
 23/Unknown - 21s 925ms/step - loss: 2.5342 - acc: 0.3750 - mIoU: 0.0561
 24/Unknown - 22s 902ms/step - loss: 2.5147 - acc: 0.3862 - mIoU: 0.0577
 25/Unknown - 22s 880ms/step - loss: 2.4969 - acc: 0.3958 - mIoU: 0.0589
 26/Unknown - 22s 861ms/step - loss: 2.4781 - acc: 0.4054 - mIoU: 0.0602
 27/Unknown - 23s 843ms/step - loss: 2.4590 - acc: 0.4152 - mIoU: 0.0617
 28/Unknown - 23s 826ms/step - loss: 2.4409 - acc: 0.4242 - mIoU: 0.0630
 29/Unknown - 23s 810ms/step - loss: 2.4240 - acc: 0.4320 - mIoU: 0.0640`