Lotayou / everybody_dance_now_pytorch

A PyTorch Implementation of "Everybody Dance Now" from Berkeley AI lab.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

compute_coordinates.py input error

SinanGncgl opened this issue · comments

When i prepare the dataset exactly what instructions say, and run the command;

python ./pose_estimator/compute_coordinates.py

It gives me the following error.

Traceback (most recent call last): File "pose_estimator/compute_coordinates_for_video.py", line 258, in <module> cord = cordinates_from_image_file(img, model=model) File "pose_estimator/compute_coordinates_for_video.py", line 188, in cordinates_from_image_file output1, output2 = model.predict(imageToTest_padded) File "/Users/sinangencoglu/anaconda3/envs/sinandl/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py", line 1096, in predict x, check_steps=True, steps_name='steps', steps=steps) File "/Users/sinangencoglu/anaconda3/envs/sinandl/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py", line 2382, in _standardize_user_data exception_prefix='input') File "/Users/sinangencoglu/anaconda3/envs/sinandl/lib/python3.7/site-packages/tensorflow/python/keras/engine/training_utils.py", line 362, in standardize_input_data ' but got array with shape ' + str(data_shape)) ValueError: Error when checking input: expected input_5 to have shape (None, None, 3) but got array with shape (184, 103, 4)

Thank you.

Seems that your video frames contain 4 channels, removing the alpha channel and use RGB input should work.