SOE-Net

• This ReadMe file explains how to use the code for trainable SOE-Net.

• This repo contains all necessary functions to build SOE-Net.

Requirements:

• This code has been tested under the following environment settings:
  • cuda 9.0
  • cudnn 7.1
  • tensorflow 1.8
  • opencv 3.4

Relevant scripts:

• SOE_Net_model_full.py
  • All necessary functions to build a trainable SOE_Net are contained in this script.
• init_SOE_Net.py
  • All functions to initialize SOE_Net's building blocks are here.
• input_data.py
  • All functions used in the input pipeline.
• util.py
  • functions to visualize videos and results.
• SOE_MSOE_SO_TEST.py
  • full fledged main code with examples to build and extract SOE, MSOE, SO, SOE-NET, MSOE-NET and SO-NET features. It needs as input:
  1. a path to a datasets folder
  2. a dataset folder
  3. a sample video to be used for testing

Citation

• if you use this code in your research please cite the corresponding paper:

@InProceedings{Hadji2017,
author={I. Hadji and R. P. Wildes},
title={A Spatiotemporal Oriented Energy Network for Dynamic Texture Recognition},
booktitle={ICCV}
year={2017}
}