/pytorch-action-recognition-toy

A toy LSTM net for action recognition using IMU sensor data, implemented in PyTorch. For the Hand-on tutorial as a TA in NAIST Spring Seminar 2019.

Primary LanguagePythonMIT LicenseMIT

Action Recognition Toy Project

This is a practice program for the Spring Seminar 2019 in the Robotics Lab of NAIST.

Environment

  • macOS Mojave 10.14.3
  • Python 3.6.8 (Anaconda 4.6.4)
  • PyTorch 1.0.0
  • TensorboardX 1.6

Usage

Run python train.py.

After running train.py, a model.pth file should have been created. Run test.py to test the model you have trained.