rfbr/pdr_with_deep_learning

pdr_with_deep_learning's Issues

Improper nesting of variable_scope

Hello, I tried to run the code but encountered several problems.

1. When using the listed environment (TensorFlow1.14.0, Cuda 10.0,.etc), I met an error (TensorFlow1.15.0 also has the same error):
   "OpHint dynamic rnn only supports control flow v2."
2. I then checked that "tf.compat.v1.lite.experimental.nn.dynamic_rnn" is available for TensorFlow 2.4.0, so I upgraded to TensorFlow 2.0.0 and 2.4.0, I met another error:
   "could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: no such file or directory"
3. I then installed cudnn 8.0 and then I met the error:
   "Exception RuntimeError occurred, arguments: ('Improper nesting of variable_scope.',)"
4. Besides, I tried CUDA10.0, CUDA10.1, CUDA11.0 and got the same results. Also, I tried to install from pip and conda and got the same results. FYI. Could you please help? Many thanks!

Training data

Hi，I want to collect data on a new device. How can I get the training data as the example?Is there an APK that can be used directly？
THX!

Parameters to consult

Hello, I am learning your project and found that the convergence was not ideal when training the model. I would like to ask you about the parameters you set at that time, such as the values of lookback size, step size and Batch size, thank you.

Questions about dataset

Hi, I have checked your dataset, and there are some questions.
1）The data acquired from ARcore, how to decide the starting point (i.e., position_x =0, position_y=0, position_z=0) in each experiment?
2）What does 'lookback' mean? You use ''acc'' and ''gyr'' of a period, to map ''position'' and ''quaternion'' in specified time point?