We are pleased to announce a dedicated Open Dataset Track at ETRA 2026. This track aims to recognize and promote high-quality datasets, benchmarks, and data-centric resources that advance the fields of eye tracking and machine learning. We invite submissions that describe the creation, curation, analysis, or benchmarking of datasets relevant to eye tracking research and applications.

We welcome submissions on (but not limited to):

• New datasets for eye tracking, gaze estimation, or related modalities (e.g., pupil dynamics, blink detection, video-based or sensor-based eye tracking, multimodal datasets).
• Data generators, simulation environments, or synthetic data resources relevant to eye tracking and human-computer interaction.
• Systematic analyses or audits of existing datasets, including coverage, bias, quality, or domain shifts.
• Tools, protocols, or best practices for dataset collection, annotation, documentation, and sharing.
• Studies on data quality, bias, privacy, fairness, or reproducibility in eye tracking datasets and pipelines.
• Meta-analyses, surveys, and comparative studies of datasets for specific applications (e.g., AR/VR, automotive, clinical, education, accessibility).

• Paper Format: Up to 6 pages (excluding references) in the standard ACM format.
• Review Model: Single-blind review; author names and affiliations should be included.
• Dataset Availability: Submissions must include a stable, accessible link to the dataset and, if applicable, code for accessing or using the data. Recommended platforms include Zenodo, Dataverse, Hugging Face, OpenML, or institutional repositories.
• Metadata: Authors are encouraged to provide machine-readable metadata (e.g., Croissant or similar formats), including modality descriptions, acquisition protocols, annotation schemas, licensing, and versioning.
• Supplementary Material: Additional documentation, sample code, data cards, or appendices may be linked from the paper (not included in the page limit).
• Ethics & Privacy: All submissions must include a 2-3 sentence Privacy and Ethics Statement addressing potential societal impacts, privacy and consent, fairness, and data-sharing considerations (see template below).
• Baseline Results: If applicable, provide baseline experiments or analyses to demonstrate dataset utility (e.g., benchmark tasks, reference pipelines, evaluation metrics).
• Licensing: Clearly state the dataset license and access conditions (e.g., CC BY, CC BY-NC, custom academic license), including any restrictions on redistribution or commercial use.
• Access for Reviewers: Datasets and code must be accessible to reviewers at submission time (e.g., public links or confidential reviewer access).

Distinct Contribution:
Dataset Track papers must present a distinct contribution and cannot duplicate the main content of an accepted or concurrent full/short paper. If your dataset is already described in a full/short paper, do not submit an identical version to the Dataset Track.

Accepted approaches:

• Full Paper with Dataset: The dataset is one of several contributions in a full paper. The Dataset Track paper may reference the full paper, but must add distinct dataset-focused documentation (e.g., expanded data card, metadata schema, licensing, access instructions).
• Dedicated Dataset Track Paper: Focused documentation and release of a new dataset with clear evidence of utility and completeness (up to 6 pages).
• Meta-Level Dataset Paper: Analysis, benchmarking, or documentation of dataset practices, including surveys, audits, and comparative studies.

Data Reuse and Updates: Submissions describing updates or extensions to existing datasets should clearly detail version changes, new content, and added value (e.g., increased diversity, improved annotations, new tasks).

All submissions will be reviewed for originality, relevance, and compliance with data availability requirements.

• Datasets and code must be accessible at submission time.
• Submissions involving controlled-access data must document the access process (including reviewer access), consent conditions, and any usage restrictions.
• Authors should detail quality assurance measures (e.g., inter-annotator agreement, calibration protocols, missing data handling) and documentation completeness (e.g., data cards, license text, metadata).

Submissions will be evaluated based on:

• Originality and scholarly value of the dataset/resource.
• Usefulness and relevance to the eye tracking community and related machine learning applications.
• Quality, completeness, and clarity of documentation, including metadata and data cards.
• Accessibility and reproducibility, including baseline results and code availability.
• Ethical, privacy, and fairness considerations, including consent, de-identification, and potential societal impacts.

Accepted Open Dataset Track papers will be published in the ETRA 2026 proceedings and included in the ACM Digital Library. Authors will be invited to present their datasets during the main conference, either as oral presentations or posters.

As part of ACM's Open Access initiative, all accepted papers will be published Open Access (OA). Please review the APC policy and possible waivers as described in the main call for papers.

This dataset was collected and shared in accordance with applicable ethical guidelines and regulatory requirements (including informed consent where appropriate). We assessed potential risks related to privacy, fairness, and misuse; data are de-identified or controlled as needed, and licensing and access conditions are designed to mitigate risks while supporting reproducible research.

Authors may adapt and expand this statement to reflect specific consent frameworks, IRB/ethics board approvals, data minimization, anonymization, and access controls.

Paper submission	March 5, 2026
Review and Notifications	March 23, 2026
Camera-ready deadline	March 30, 2026

Please direct any questions regarding the Open Dataset Track to the Dataset Track Chairs at dataset.etra2026@acm.org.