search

hx173149 / C3D-tensorflow

C3D is a modified version of BVLC tensorflow to support 3D ConvNets.
MIT License
588 stars 265 forks source link

Issue in Training!!!! #99

Open abhisheksurya578 opened 5 years ago

abhisheksurya578 commented 5 years ago

Hi @gudongfeng @hx173149 @chuckcho @cc786537662

ValueError: Cannot feed value of shape (40, 0) for Tensor u'Placeholder:0', which has shape '(?, 16, 128, 128, 3)'

Can you please suggest what could be the problem?

abhisheksurya578 commented 5 years ago

Traceback (most recent call last): File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 262, in tf.app.run() File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 48, in run _sys.exit(main(_sys.argv[:1] + flags_passthrough)) File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 259, in main run_training() File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 224, in run_training labels_placeholder: train_labels File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 789, in run run_metadata_ptr) File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 975, in _run % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape()))) ValueError: Cannot feed value of shape (20, 0) for Tensor 'Placeholder:0', which has shape '(20, 16, 112, 112, 3)' help me to resolve this

Shiny128 commented 5 years ago

Hello, Please Check the following issue , helped me solve a similar error Duplicate of #47

abhisheksurya578 commented 5 years ago

Hello @Shiny128 @gudongfeng @hx173149 @chuckcho @highway007

I want to train the model with new Dataset .

How to create crop_mean.npy file for new dataset?

What are the changes to be made to C3D model?(Removing the final layers)

I wanna detect the object in the video as well how to do this?

Can anyone please guide me...

qizhiliu0639 commented 4 years ago

Traceback (most recent call last): File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 262, in tf.app.run() File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 48, in run _sys.exit(main(_sys.argv[:1] + flags_passthrough)) File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 259, in main run_training() File "/home/ltroot/data1/C3D-tensorflow-master/train_c3d_ucf101.py", line 224, in run_training labels_placeholder: train_labels File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 789, in run run_metadata_ptr) File "/home/ltroot/PycharmProjects/C3D/venv/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 975, in _run % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape()))) ValueError: Cannot feed value of shape (20, 0) for Tensor 'Placeholder:0', which has shape '(20, 16, 112, 112, 3)' help me to resolve this

I have met this problem before, it might be the input_data.py's problem, it didn't read your file. You maybe need to do a little change to the os.walk()