gazescrnn

Official implementation for 'GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network'

https://github.com/stijngroenen/gazescrnn

Science Score: 67.0%

This score indicates how likely this project is to be science-related based on various indicators:

✓
CITATION.cff file
Found CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
✓
DOI references
Found 2 DOI reference(s) in README
✓
Academic publication links
Links to: arxiv.org
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (10.3%) to scientific vocabulary

Last synced: 6 months ago · JSON representation ·

Repository

Official implementation for 'GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network'

Basic Info

Host: GitHub
Owner: StijnGroenen
License: mit
Language: Python
Default Branch: main
Homepage: https://doi.org/10.48550/arXiv.2503.16012
Size: 29.3 KB

Statistics

Stars: 1
Watchers: 1
Forks: 0
Open Issues: 0
Releases: 0

Created over 1 year ago · Last pushed 12 months ago

Metadata Files

Readme License Citation

GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network

This repository contains the official implementation for GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network.

GazeSCRNN is a spiking convolutional recurrent neural network designed for event-based near-eye gaze tracking. This repository contains the official implementation of the model, training scripts, and evaluation metrics.

Reference

If you find our paper or this repository helpful, please consider citing:

```bibtex @misc{groenen2025gazescrnn, title={GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network}, author={Stijn Groenen and Marzieh Hassanshahi Varposhti and Mahyar Shahsavari}, year={2025}, eprint={2503.16012}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2503.16012}, }

```

Requirements

To install the required dependencies, run:

bash pip install -r requirements.txt

Besides the Python dependencies, training and testing the GazeSCRNN models requires the EV-Eye dataset to be present in the EVEyedataset directory. The EV-Eye dataset can be obtained by following the steps here.

Training

To train the GazeSCRNN model, run the train.py script with the desired parameters. For example:

bash python train.py Experiment1 --data_preload --gpus 0 --fptt

Alternatively, you can train the GazeSCRNN model with one of the predefined configurations. For example:

bash xargs python train.py --gpus 0 < configs/GazeSCRNN-Events300-FPTT-Backprop8.txt

Testing

To test a checkpoint of the GazeSCRNN model, run the test.py script with the desired parameters:

bash python test.py <experiment_name> <path_to_checkpoint_file> --gpus <gpu_id>

This will output the evaluation metrics such as Mean Angle Error (MAE), Mean Pupil Error (MPE), and Mean Firing Rate (MFR).

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Owner

Login: StijnGroenen
Kind: user
Location: The Netherlands

Website: https://stijngroenen.nl
Repositories: 2
Profile: https://github.com/StijnGroenen

Citation (CITATION.cff)

cff-version: 1.2.0
title: "GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network"
message: "If you use this software, please cite it using the metadata from this file."
authors:
    - given-names: Stijn
      family-names: Groenen
      affiliation: Radboud University
      orcid: "https://orcid.org/0009-0003-7042-2954"
    - given-names: Marzieh
      family-names: Hassanshahi Varposhti
      orcid: "https://orcid.org/0009-0005-7958-2051"
      affiliation: "Donders Institute for Brain, Cognition, and Behaviour"
    - given-names: Mahyar
      family-names: Shahsavari
      affiliation: "Donders Institute for Brain, Cognition, and Behaviour"
      orcid: "https://orcid.org/0000-0001-7703-6835"
identifiers:
    - type: doi
      value: 10.48550/arXiv.2503.16012
      description: arXiv
repository-code: "https://github.com/StijnGroenen/GazeSCRNN"
license: MIT
preferred-citation:
    authors:
        - given-names: Stijn
          family-names: Groenen
          affiliation: Radboud University
          orcid: "https://orcid.org/0009-0003-7042-2954"
        - given-names: Marzieh
          family-names: Hassanshahi Varposhti
          orcid: "https://orcid.org/0009-0005-7958-2051"
          affiliation: "Donders Institute for Brain, Cognition, and Behaviour"
        - given-names: Mahyar
          family-names: Shahsavari
          affiliation: "Donders Institute for Brain, Cognition, and Behaviour"
          orcid: "https://orcid.org/0000-0001-7703-6835"
    title: "GazeSCRNN: Event-based Near-eye Gaze Tracking using a Spiking Neural Network"
    doi: 10.48550/arXiv.2503.16012
    type: article
    journal: "arXiv preprint arXiv:2503.16012"
    year: 2025

GitHub Events

Total

Watch event: 1
Push event: 1
Create event: 2

Last Year

Watch event: 1
Push event: 1
Create event: 2

Dependencies

requirements.txt pypi

h5py ==3.10.0
lightning ==2.2.1
numpy ==1.23.5
pandas ==2.2.3
tensorboard ==2.16.2
tonic ==1.4.3
torch ==2.4.1
torchvision ==0.19.1
tqdm ==4.66.2

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Open Source Science