search

VIPL-Audio-Visual-Speech-Understanding / learn-an-effective-lip-reading-model-without-pains

The PyTorch Code and Model In "Learn an Effective Lip Reading Model without Pains", (https://arxiv.org/abs/2011.07557), which reaches the state-of-art performance in LRW-1000 dataset.
152 stars 37 forks source link

RuntimeWarning: invalid value encountered in double_scalars #9

Closed PatrickPrakash closed 3 years ago

PatrickPrakash commented 3 years ago

Hey there, I tried to run your program and get this warning. Can you help me out. I'm just starting deeplearning 😊

load weights
loaded params/tot params:151/151
miss matched params: []
Start Testing, Data Length: 0
start testing
main_visual.py:155: RuntimeWarning: Mean of empty slice.
  acc = float(np.array(v_acc).reshape(-1).mean())
/home/ubuntu/anaconda3/envs/learn_lip/lib/python3.8/site-packages/numpy/core/_methods.py:188: RuntimeWarning: invalid value encountered in double_scalars
  ret = ret.dtype.type(ret / rcount)
acc=nan
Fengdalu commented 3 years ago

The program prints Start Testing, Data Length: 0, which means it does not read any testing data. You may specific data path to read data properly.

PatrickPrakash commented 3 years ago

I didn't find any dataset path specific argument in main python file. I link the lrw dataset to the root as per documentation and run the argument as per documentation. Can you please help me out how can I specify the test dataset?Thanks

Fengdalu commented 3 years ago

Can you find lrw_roi_80_116_175_211_npy_gray_pkl_jpeg in your project folder after run scripts/prepare_lrw.py ?

PatrickPrakash commented 3 years ago

Yeah , I saw that

PatrickPrakash commented 3 years ago

But nothing is inside the folder

PatrickPrakash commented 3 years ago

The issue was solved by changing the name of dataset which was "lrw_mp4" but actually was "lipread_mp4". It will be great if you could share your social media handle or email so that i can contact you for further questions.

Fengdalu commented 3 years ago

You can contact me via fengdalu@gmail.com or dalu.feng@vipl.ict.ac.cn. Moreover, sharing your problem in Github issues is recommended for further disscutions.