Implementation of GrNet [1], a deep network on Grassmann manifolds.

• Python 3.6+
• SciPy
• NumPy
• TensorFlow 2.0+
• TensorFlow RiemOpt

Configure gcloud to use Python 3:

gcloud config set ml_engine/local_python /usr/bin/python3

Train GrNet locally on the Acted Facial Expression in Wild [2] dataset:

gcloud ai-platform local train \
       --module-name grnet.task \
       --package-path . \
       -- \
       --data-dir data
       --job-dir ckpt

1. Huang, Zhiwu, Jiqing Wu, and Luc Van Gool. "Building Deep Networks on Grassmann Manifolds." AAAI. AAAI Press, 2018.
2. Dhall, Abhinav, et al. "Acted facial expressions in the wild database." Australian National University, Canberra, Australia, Technical Report TR-CS-11 2 (2011): 1.