docs/index.md


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51

---
title: What is ArGaze?
---

# Develop post- or real-time gaze analysis applications

**Useful links**: [Installation](installation.md) | [Source Repository](https://gitpub.recherche.enac.fr/argaze) | [Issue Tracker](https://git.recherche.enac.fr/projects/argaze/issues) | [Contact](mailto:argaze-contact@recherche.enac.fr)

**ArGaze** is an open and flexible Python software library designed to provide a unified and modular approach to gaze analysis or gaze interaction.  
  
By offering a wide array of gaze metrics and supporting easy extension to incorporate additional metrics, **ArGaze** empowers researchers and practitioners to explore novel analytical approaches efficiently.

![ArGaze pipeline](img/argaze_pipeline.png)

## Eye tracking context

**ArGaze** facilitates the integration of both **screen-based and head-mounted** eye tracking systems for **live data capture and afterward data playback**.

[Learn how to handle various eye tracking context by reading the dedicated user guide section](./user_guide/eye_tracking_context/introduction.md).

## Gaze analysis pipeline

Once incoming eye tracking data available, **ArGaze** provides an extensible modules library, allowing to select application-specific algorithms at each pipeline step:

* **Fixation/Saccade identification**: dispersion threshold identification, velocity threshold identification, etc.  
* **Area Of Interest (AOI) matching**: focus point inside, deviation circle coverage, etc.
* **Scan path analysis**: transition matrix, entropy, explore/exploit ratio, etc.

All those gaze analysis features can be used with any screen-based eye tracker devices.

[Learn how to build gaze analysis pipelines for various use cases by reading the dedicated user guide section](./user_guide/gaze_analysis_pipeline/introduction.md).

## Augmented reality based on ArUco marker pipeline

Things goes harder when gaze data comes from head-mounted eye tracker devices. That's why **ArGaze** provides **Augmented Reality (AR)** support to map **Areas Of Interest (AOI)** on [OpenCV ArUco markers](https://www.sciencedirect.com/science/article/abs/pii/S0031320314000235).

![ArUco pipeline axis](img/aruco_pipeline_axis.png)

This ArUco marker pipeline can be combined with any wearable eye tracking device Python library, like Tobii or Pupil glasses.

[Learn how to build ArUco marker pipelines for various use cases by reading the dedicated user guide section](./user_guide/aruco_marker_pipeline/introduction.md).

!!! note
   
      *ArUco marker pipeline is greatly inspired by [Andrew T. Duchowski, Vsevolod Peysakhovich and Krzysztof Krejtz article](https://git.recherche.enac.fr/attachments/download/1990/Using_Pose_Estimation_to_Map_Gaze_to_Detected_Fidu.pdf) about using pose estimation to map gaze to detected fiducial markers.*

## Demonstration

![type:video](https://achil.recherche.enac.fr/videos/argaze_features.mp4)

[Test **ArGaze** by reading the dedicated user guide section](./user_guide/utils/demonstrations_scripts.md).