https://github.com/VRatPolito/LET-VR
Tip revision: 3765fd6531da76fd25f4449f3ca7e9081c8c6877 authored by F. Gabriele Pratticò on 26 October 2022, 16:16:31 UTC
Update README.md
Update README.md
Tip revision: 3765fd6
README.md
<p align="center" width="100%">
<img width="350" src="https://i.postimg.cc/JhCd9T52/giffone-2c.gif">
</p>
# Locomotion Evaluation Testbed VR (OpenXR - Unity v2020LTS)
> **Locomotion Evaluation Testbed VR** (or **LET-VR**) is a research project aimed at supporting a comprehensive
> comparison of locomotion techniques (for immersive VR) using a provided evaluation testbed.
> The testbed application is complemented by
> - an experimental protocol for collecting objective and subjective measures
> - a scoring system able to **rank** locomotion techniques based on a **weighted** set of requirements
>
> **LET-VR** helps to select the best suitable locomotion technique to adopt in a given VR application scenario.
> To dig more into the details about the testbed design, you can refer to the paper (see [**Citation**](#citation))
<p align="center" width="100%">
<img width="600" src="https://i.postimg.cc/FK4923pp/architettura.png">
</p>
## Table of contents
* [Introduction](#introduction)
* [Videos](#videos)
* [Experimental Material](#experimental-material)
* [Builds and Building](#builds)
* [Configuration](#configuration)
* [Custom Locomotion Technique](#implementing-a-custom-locomotion-technique)
* [Forking Policies](#forking-policies)
* [Citation](#citation)
* [Contacts](#contact)
* [License](#license)
* [Acknowledgements](#acknowledgements)
## Introduction
For an overview of the project have a look at the video here below presented at the\
*ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games* (i3D 2021)\
\
[](https://youtu.be/si1omc3T9W4?t=7600 "Video Title")
## Videos
Some videos showing tasks execution with different locomotion techniques (Arm-Swinging, Walking-In-Place, Cyberith
Virtualizer, and Joystick) for every scenarios (training included) are available at this [**Link**](http://tiny.cc/8uxlsz)
## Experimental Material
Additional material to support the user-study can be found in the [**Experimental Material/**](Experimental%20Material/)
folder. In particular:
- [**Administrator Script.odt**](Experimental%20Material/Administrator%20Script.odt): Script supporting the test
administrator in providing information to the participant
- [**( R ) WDB.xlsx**](Experimental%20Material/(R)WDB.xlsx): Example of WDB (plus Raw data) with 4 techniques comparison
- [**Questionnaire.ods**](Experimental%20Material/Questionnaire/Questionnaire.ods): The full questionnaire in a
spreadsheet format
- [**Questionnaire_printable.pdf**](Experimental%20Material/Questionnaire/Questionnaire_printable.pdf): The **same**
full questionnaire in a printable format
## Builds
The latest version of the executables are already available for the download in [**Release**](https://github.com/VRatPolito/LET-VR/releases)
The application targets only **Windows 10/11** and can be deployed to any VR system compatible with the **OpenXR** API,
although minimum modifications may be necessary to support non-tested hand controllers
- The builds have been tested with **HTC Vive/Pro** w/ controllers (+ Vive Trackers for the WIP) and **Oculus Quest 2 (
via AirLink)** w/ touch controllers.
- Please note that even tough any **OpenXR Runtime can be used**, to correctly load the 3D models of the controllers it
is required **SteamVR** (any version) being installed on the host machine.
### Build *LET-VR* from scratch
This is a new version of the project revamped to work with newer Unity versions and the OpenXR backend. If you are
looking for the version published in [**Citation**](#citation), have a look at the former [**master_public**](https://github.com/VRatPolito/LET-VR/tree/master_public) branch.
Instructions to compile the project:
#### Infos
The project was revamped and tested with [**Unity 2020.3.x (
LTS)**](https://unity3d.com/unity/qa/lts-releases?version=2020.3)
- NOTE: this project version uses **OpenXR** and the new **Unity Input System** integrated into UnityXR. However, for **
backward compatibility** with the Logging system, and consistency with the former version of the project the inputs
and interaction with the virtual environment are **NOT** managed using the novel UnityXR InteractionToolkit. Have a
look at the class `InputManagement.cs` responsible for the wrapping.
The list of package dependencies is in the [**manifest file**](UnityProject/Packages/manifest.json) and will be
automatically managed by the Unity editor. You will need also [**Blender**](https://www.blender.org/download/) to be
installed (v2.8+), to have Unity correctly load the blend files in the project.
#### Prepare to Build
1. In Unity, open an existing project and select the [**UnityProject/**](UnityProject/) folder
1. If package errors are reported, press Continue, open the Package Manager (Window -> Package Manager) and try
updating the involved packages, after that restart the project
1. Scene files for each scenario are placed inside the [**UnityProject/Assets/Scenes/**](UnityProject/Assets/Scenes/)
folder. Before building, open each scene file and perform the bake of the lighting (Window -> Lighting Tab ->
Generate Lighting).
1. NOTE: baking is a computationally intensive task, and the time required for completing it can vary based on the
hardware (in particular, for Scenario 2). By default, Progressive GPU Lightmapper is selected, switch back to CPU
in case of low-performance graphics adapters. Also, to pick a specific GPU device to be used for the baking
please refer to the [**official Unity manual
page**](https://docs.unity3d.com/2018.4/Documentation/Manual/GPUProgressiveLightmapper.html)
1. The *CybSDK* and *KAT_Walk SDK* plug-in required to compile with support to the *Cyberith Virtualizer/Kat Walk*,
is **NOT** included within the project source due to licensing. To support the locomotion treadmill it is required to
obtain and import the relative unity package.
1. If you are not able to get the *CybSDK/KatWalkSDK* but want to test **LET-VR** on the *Cyberith Virtualizer/Kat
Walk* you can use the compiled version in [**Release**](https://github.com/VRatPolito/LET-VR/releases)
#### Build facilities
- To Build it from scratch, use the facility in the custom menu VR@POliTO (N.B. remember to bake lighting for all the
scenarios):
- **Build All Scenarios Split**: build all scenarios in a different path with each executable including just one
scenario (e.g. *Scenario1.exe*)
- **Build Launcher (built-in scenarios)**: build all scenarios embedded into a single executable (*Launcher.exe*)
<p align="center" width="100%">
<img width="600" src="http://vr.polito.it/wp-content/uploads/2020/01/build_helpers.png">
</p>
## Configuration
The **LET-VR** testbed application can be configured with additional parameter without the need to rebuild it every
time. If using the **Launcher** mode this can be done by means of the provided configuration mask.
<p align="center" width="100%">
<img width="500" src="https://i.postimg.cc/TYLXqD4t/launcher.png">
</p>
Otherwise, is possible to manually edit the configuration file. The file is located in the *BuildData/* folder,
e.g. `..\LET-VR\Launcher\Launcher_Data\BuildData\config.txt`
This is an example of a `config.txt` file:
#Select scenario to load [Launcher only] {-1 = SelectWithGUI, 0 = Training, 1 = Straight Movements, 2 = Direction Control, 3 = Decoupled Movements, 4 = Agility, 5 = Interaction with objects}
Scenario=-1
#Participant ID (can be a string) [max lenght 15]
ParticipantId=56
#SetLocomotion Technique among {ArmSwing, WalkInPlace, CVirtualizer, Joystick}
LocomotionTechnique=ArmSwing
#Use true to automatically freeze the player at the start/end of each task
AutoFreeze=true
#### Calibration Data
At the first run, a calibration file `calibrationData.pgd` with the participant data will be created in the User's
document folder, e.g. `..\FooUser\Documents\LET_VR`. The file can be automatically populated by using the *Tutorial
Scenario* or manually by editing the json file.
```json
{
"_controllerDistance": 1.5,
"_headHeight": 1.7
}
```
## Implementing a *Custom* Locomotion Technique
We have not fully documented yet how to integrate a novel technique into the project but we have plans to do. In the
meanwhile you can start by looking at the `LocomotionManager.cs` which should be enough self-explaining, and at the **Prefab** of **Joystick Controller** (and respective `JoystickMovement.cs`) in a given scene, and explore the hierarchy.
**N.B.** It is important for the logging system to have your technique to work with the `CharacterController`
## Forking Policies
Please contact [F. Gabriele Prattico\`](mailto:filippogabriele.prattico@polito.it?subject=[GitHub]%20LET-VR) **BEFORE**
forking the Project
## Citation
Please cite this paper in your publications if it helps your research.
@article{letvr,
author = {Cannav{\`o}, Alberto and Calandra, Davide and Prattic{\`o}, Filippo Gabriele and Gatteschi, Valentina and Lamberti, Fabrizio},
journal = {IEEE Transactions on Visualization and Computer Graphics},
title = {An Evaluation Testbed for Locomotion in Virtual Reality},
pages={1871-1889},
year = {2020},
doi = {10.1109/TVCG.2020.3032440}
}
The **LET-VR** design is detailed in:
- *An Evaluation Testbed for Locomotion in Virtual Reality*
- [**IEEE TVCG**](https://dx.doi.org/10.1109/TVCG.2020.3032440)
- [**ArXiv**](https://arxiv.org/abs/2010.10178)
- Have a look at the introductory video in [**Introduction**](#introduction)
## Contact
Maintained by [F. Gabriele Prattico\`](mailto:filippogabriele.prattico@polito.it?subject=[GitHub]%20LET-VR) - feel free
to contact me!
## License
Experimental material and Unity project are licensed under
[](https://creativecommons.org/licenses/by/4.0/)
## Acknowledgements
<p align="center" width="100%">
<img width="800" src="https://vr.polito.it/wp-content/uploads/2021/09/logo_intero_vr_polito_novel.png">
</p>