This repository is the official implementation of [Re] IDOL: Inertial Deep Orientation-Estimation and Localization. The code is being prepared for submission to: (https://paperswithcode.com/rc2021)ML Reproducibility Challenge 2021 Fall Edition, and a course project in CISC 867 Deep Learning, Queen's University.
📋 Optional: include a graphic explaining your approach/main result, bibtex entry, link to demos, blog posts and tutorials
Optional Dependencies piptools - used to modify requirements.txt without dependency hassles.
pip install pip-tools
Steps:
pip install pip-tools
to use the command below.pip-compile
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
To setup the datasets: a. Create a folder called datasets. b. Create another folder within datasets called csvs. You should have datasets/csvs as part of your folder structure c. Download and extract the datasets from here. Extract each building into datasets.
To install tensorflow graphics
a. Run git clone https://github.com/tensorflow/graphics.git
.
b. cd
to the directory where you cloned tensorflow graphics.
c. Run python -m venv .venv
d. Run source .venv/bin/activate
(Bash), .\.venv\Scripts\activate.ps1
(Windows Powershell) or
.\.venv\Scripts\activate.bat
(Windows Cmd)
e. Run pip install wheel
f. Run python setup.py bdist_wheel
g. cd
to Re-IDOL's location.
h. Run pip install /path/to/tensorflow-graphics-location/dist/tensorflow_graphics-2021.12.11-py3-none-any.whl
To train the model(s) in the paper, run this command:
python main.py train_orient --option=<option number 1-3>
python main.py train_pos --option=<option number 1-3>
To test the model(s) in the paper, run this command:
python main.py test_orient --option=<option number 1-3>
python main.py test_pos --option=<option number 1-3>
You can find pretrained models in the directory: pretrained/Buildings"<number 1-3>"/OrientNet
Our model achieves the following performance on the known set (training) of buildings (1-3)/ the unknown set (1-3):
Model name | Building 1 | Building 2 | Building 3 |
---|---|---|---|
OrientNet (rad) | x / test | x / test | tbd |
PosNet (meter) | y / test | y / test | tbd |
📋 Include a table of results from your paper, and link back to the leaderboard for clarity and context. If your main result is a figure, include that figure and link to the command or notebook to reproduce it.
Uses Apache License, see LICENSE for more details.