Non-invasive vision-based personal comfort model using thermographic images and deep learning

This is the official repository for Non-invasive vision-based personal comfort model using thermographic images and deep learning, our paper published in the Journal of Automation in Construction.

The paper can be found here

Prerequisites

Create a conda environment and install the following:

1. Tensorflow
2. Matplotlib
3. Scikit-learn `

Data Preparation

Your dataset should have the following structure:

- 3_TSV_base_dir/
  - participant 1/
  - participant 2/
  ...
      - Cool
      - Neutral
      - Warm

The same file structure should be used for 7 TSV.

Training

Run the following script for 7 TSV training python main.py --base_dir ./base_directory --num_classes 7

Run the following script for 3 TSV training python main.py --base_dir ./base_directory --num_classes 3