Skip to content

Commit

Permalink
@N-M-T 's review and edits
Browse files Browse the repository at this point in the history
  • Loading branch information
N-M-T committed Oct 2, 2023
1 parent cfd8ab9 commit 4af7afd
Showing 1 changed file with 22 additions and 9 deletions.
31 changes: 22 additions & 9 deletions src/alpha-lab/scanpath-rim.md
Original file line number Diff line number Diff line change
Expand Up @@ -17,25 +17,38 @@ tags: [Pupil Invisible, Neon, Cloud]
Picture this: Build and customise scanpath visualisations with your Reference Image Mapper exports!

Check warning on line 17 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (customise)

Check warning on line 17 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (scanpath)

Check warning on line 17 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (visualisations)
:::

## Unlocking visual exploration with scanpaths
A scanpath is a graphical representation of an individual's gaze. It shows the sequence of fixations (pauses in gaze), and rapid eye movements made between fixations (saccades). Scanpaths offer a glimpse into how the observer has focused their attention on different aspects of the scene, which is a valuable tool for understanding a person's visual attention and perception.
## Visualising gaze exploration with scanpaths

Check warning on line 20 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (Visualising)

Check warning on line 20 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (scanpaths)
Scanpaths are graphical representations of gaze over time. As shown in the video above, they contain fixation locations

Check warning on line 21 in src/alpha-lab/scanpath-rim.md

View workflow job for this annotation

GitHub Actions / ✍️ Check spelling

Unknown word (Scanpaths)
(blue numbered circles) and saccades made between fixations (blue connecting lines). The size of the circles reflect
fixation durations, with longer fixations corresponding to bigger circles. The length of the connecting lines reflect
saccadic amplitudes, with larger gaze shifts corresponding to longer lines. Scanpaths offer a glimpse into how the
observer has focused their attention on different aspects of the scene, which is a valuable tool for understanding a
person's visual attention and perception.

In this guide, we will show you how to generate both static and dynamic scanpath visualisations using your Reference Image Mapper exported data, like in the video above.
In this guide, we will show you how to generate both static and dynamic scanpath visualisations using your Reference
Image Mapper exported data.

::: tip
Before continuing, ensure you are familiar with the [Reference Image Mapper](/enrichments/reference-image-mapper)
enrichment. Check out [this explainer video](https://www.youtube.com/watch?v=ygqzQEzUIS4&t=56s) for reference.
:::

## Building the visualisations in an offline context
The [Reference Image Mapper](/enrichments/reference-image-mapper) available in Pupil Cloud is a tool that maps gaze onto 2D images and can subsequently generate heatmaps. However, it currently does not support the production of scanpath visualizations. Since scanpaths provide a useful characterization of *where*, *when*, and *how long* attention was focused on various elements, we developed a script that enables you to generate both static and dynamic scanpaths using your Reference Image Mapper data exported from Pupil Cloud.
## Extending current tools
The [Reference Image Mapper](/enrichments/reference-image-mapper) enrichment available in Pupil Cloud is a tool that maps gaze onto
2D images and can subsequently generate heatmaps. However, it currently does not support the production of scanpath visualizations.
Thus, we chose to develop a script that shows you how to build your own scanpaths using Reference Image Mapped data.


## Steps
1. Run a [Reference Image Mapper](https://docs.pupil-labs.com/enrichments/reference-image-mapper/) enrichment and download the results
2. Download [this](https://gist.github.com/elepl94/9f669c4d81e455cf2095957831219664) gist and follow the instructions in the [readme](https://gist.github.com/elepl94/9f669c4d81e455cf2095957831219664#file-readme-md)
1. Run a [Reference Image Mapper enrichment](https://docs.pupil-labs.com/enrichments/reference-image-mapper/) and download the results
2. Download [this script](https://gist.github.com/elepl94/9f669c4d81e455cf2095957831219664) and follow the [installation instructions](https://gist.github.com/elepl94/9f669c4d81e455cf2095957831219664#installation)

## Final results
After the script has completed its execution, you'll find the resulting scanpath visualizations stored in a newly created sub-folder named "scanpath." For each participant, you will obtain a reference image with the scanpath superimposed on it, along with a video that illustrates gaze behavior on the reference image, featuring a dynamic scanpath overlay. Furthermore, an aggregated visualization, combining all participants' scanpaths, will be at your disposal, providing the opportunity for more comprehensive and in-depth analyses.
## Review the scanpaths
After the script has completed its execution, you'll find the resulting scanpath visualizations stored in a newly created
sub-folder named "scanpath." For each participant, you will obtain a reference image with the scanpath superimposed on it.
You will also find a video featuring a dynamic scanpath overlay. Finally, if you had multiple participants, an aggregated
visualization combining all participants' scanpaths will be available, enabling a more comprehensive overview of the subjects'
gaze behavior.

<div style="display: flex; justify-content: space-between;">
<div style="flex: 1; margin-right: 10px;">
Expand Down

0 comments on commit 4af7afd

Please sign in to comment.