Skip to content

Latest commit

 

History

History
19 lines (14 loc) · 1.08 KB

Week2_update.md

File metadata and controls

19 lines (14 loc) · 1.08 KB
Raw

This Week

  1. Converted LIDAR MOCAP data to readable .skeleton format, including being padded w/ appropriate number of joints, with help from Alex & Joe.
  2. Received Wahab access, set up testing environment & accompanying Slurm scripts to run with help from Meghan.
  3. Adapted existing code to work with updated version of Numpy & PyTorch. Created pull request on original Github repo.
  4. Successfully trained and tested HCN-PyTorch model on NTURGBD data, received ~94% accuracy on testing and evaluation data.
  5. Attended DeapSECURE deep-learning seminar for 3 days. Gained more experience with Pandas, Numpy, Scikit-Learn, and working remotely on Wahab.
  6. Presented bi-weekly progress for Dr. Xin, who seemed very please.
  7. Friday, Saturday, Sunday will be devoted to running the now-working HCN-PyTorch model on our converted LIDAR MOCAP data.

Next Week

  1. Finish running HCN-PyTorch model on our data
  2. Begin learning GAN to create noisy, more difficult data.

Future

  1. Continue meeting with Dr. Khan, Alex, Joe, Meghan to help decide further progress for the project.

6/17/21 bk