[Re] IDOL: Inertial Deep Orientation-Estimation and Localization

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.

1. Quaternion Multiplication -> See here
2. Yury Petrov's Ellipsoid Fitting (Python Version) -> See here
3. Extended Kalman Filters -> See here
4. 3Blue1Brown Quaternion Explanations -> See here
5. IMUs and what they do -> See here
6. What is a random walk? -> See here

📋 Optional: include a graphic explaining your approach/main result, bibtex entry, link to demos, blog posts and tutorials

Requirements

Optional Dependencies piptools - used to modify requirements.txt without dependency hassles.

pip install pip-tools

Steps:

1. (Optional) To generate new requirements (after adding new requirement to requirements.in): Note: This requires that you install pip-tools, if you haven't installed pip-tools then please do pip install pip-tools to use the command below.

pip-compile

2. To setup virtual environment:

python -m venv .venv

3. To activate virtual environment (unix):

source .venv/bin/activate

4. To install requirements:

pip install -r requirements.txt

5. 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.

6. 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

Training

To train the model(s) in the paper, run this command:

OrientNet

python main.py train_orient --option=<option number 1-3>

PosNet

python main.py train_pos --option=<option number 1-3>

Evaluation

To test the model(s) in the paper, run this command:

OrientNet

python main.py test_orient --option=<option number 1-3>

PosNet

python main.py test_pos --option=<option number 1-3>

Pre-trained Models

You can find pretrained models in the directory: pretrained/Buildings"<number 1-3>"/OrientNet

Results

Our model achieves the following performance on the known set (training) of buildings (1-3)/ the unknown set (1-3):

Dataset

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.

Contributing

Uses Apache License, see LICENSE for more details.