Reproducible Research in Pattern Recognition (RRPR) Badge

The ensemble of practices of what we refer as reproducible research allow for reliable and trustworthy scientific advance. Science builds and organizes knowledge about refutable facts, and therefore the scientific method requires refutability, repeatability, and reproducibility of the experiments and their results.

All authors submitting their work to ICPR are encouraged to make it reproducible so other researchers can learn and build upon it. This includes adding detailed procedures in the paper that allow other researchers to re-create and validate the results and conclusions.

In this edition of ICPR we will grant a special Reproducible Research in Pattern Recognition (RRPR) Badge to accepted papers that follow the principles of reproducible research, transparency, and scientific trustworthiness, as determined by a separate evaluation by specially designated RRPR reviewers and the Reproducibility Chairs. Such a badge is a visible marker made of an icon and a small text “This work has received the ICPR 2026 Reproducible Research in Pattern Recognition (RRPR) Badge after a positive peer-review evaluation”, to be added to the camera-ready version.

To be considered for the RRPR Badge, authors must make their source code and data available for review. They must also make clear the requirements for any member of the research community to obtain access to their code and data. For example, there could be a requirement to sign a statement that the code and data will only be used for non-commercial purposes.

In the particular case where a data source is public but still it is not possible that other researchers retrieve exactly the same data, the authors can provide a sample of the subset they used in their paper. This assumes that the source data is under a license that allows reuse and share. For example, we could think of a source of live video used for object recognition and the authors providing a clip of video so others can test and validate the results.

The evaluation criteria is at the discretion of the specially-designated RRPR reviewers, who will evaluate if the paper contains all the elements allowing other researchers to obtain the same or similar results, and to reach the same conclusions as the paper. Building on the reviews provided by the RRPR reviewers, the Reproducibility Chairs will add their own meta-review to complete the evaluation.

This review will only take place for accepted papers after acceptance decisions have been made by the conference Program Committee. Because the RRPR Badge evaluation comes afterwards, the reproducibility review will have no impact on acceptance decisions.

A selection of papers will be presented at the RRPR workshop and all ICPR attendees are welcomed to participate in this workshop, for which we expect substantial discussions.

Please note that the workshops are not part of the main conference and therefore authors need to make another submission to the workshop.

This Reproducible Research in Pattern Recognition (RRPR) Badge is an acknowledgement of the effort made by authors towards cumulative and reliable research, that we hope will increase their visibility and mean a positive impact as a merit in their career.