# Scene Statistics

This repository contains the experimental data and scripts to perform data and image analysis as described in "The visibility of Eidolon distortions in THINGS and STUFF".

## Sources

Within this implementation the following work of previous projects was used.

### Algorithms
- Eidolon distortions: https://www.allpsych.uni-giessen.de/EidolonFactories/index.htm  
        [1] Jan Koenderink, Matteo Valsecchi, Andrea van Doorn, Johan Wagemans, and Karl Gegenfurtner (2017). Eidolons: Novel stimuli for vision research. Journal of Vision. 17(2). (https://doi.org/10.1167/17.2.7)
- LGN algorithm: https://github.com/irisgroen/LGNstatistics  
        [2] Iris I.A. Groen, Sennay Ghebreab, Hielke Prins, Victor A.F. Lamme, and H. Steven Scholte (2013). From image statistics to scene gist: Evoked neural activity reveals transition from low-level natural image structure to scene category. The Journal of Neuroscience. 33(48) (https://doi.org/10.1523/JNEUROSCI.3128-13.2013)

### Databases
- STUFF: [3] Filipp Schmidt, Martin N. Hebart, Alexandra C. Schmid, and Roland W. Fleming (2022). Core dimensions of human material perception. PsyArXiv (https://doi.org/10.31234/osf.io/jz8ks)
- THINGS: https://things-initiative.org  
        [4] Martin N. Hebart, Adam H. Dickter, Alexis Kidder, Wan Y. Kwok, Anna Corriveau, Catilin Van Wicklin, and Chris I. Baker (2019). THINGS: A database of 1,854 object concepts and more than 26,000 naturalistic object images. PLOS ONE. 14(10) (https://doi.org/10.1371/journal.pone.0223792)

## Stimuli

The stimuli used for this project can be found at: TODO: add stimuli some were

## Basic directory structure

- `data`: Raw and preprocessed data from experiments.
- `docs`: Documents for data collection
- `figures`: Resulting figures of the scripts.
- `publications`: TeX and pdf files of the paper.
- `results`: Outputs of analysis.
- `scripts`: Anything that is "run" as a script -- including notebook-like objects.
- `src`: Source code that can be imported into e.g. scripts. Some of the functions rely on `src` being installed as package: `pip install -e .`
- `tests` (aspirationally): Tests for code in `src`.

## Setup

### Requirements
This project provides two requirements files to build environments.
One for running the experiment (using PsychoPy) and one for performing data and distortion analysis (using numpyro etc.).

### Local setup
As the image data bases are not part of this repository it is necessary to download the images manually and add pointers to were the images are stored.

This can be done but updating the `local_device_config.json` file, in particular the paths to the image folders.