Sharing data and code to allow others to replicate research results is the ideal way to advance the field and improve multimedia systems. To reach this goal, the “Open Dataset & Software Track” at MMSys provides an opportunity for researchers and practitioners to make their work available and citable, as well as to increase the public awareness of their considerable efforts.

Those who have created a new dataset or open source software package that is relevant to the multimedia community should consider submitting it to this track. Together with the dataset or source code, authors are asked to provide a short paper describing its motivation, design, and usage, as well as discussing the way it can be useful to the community.

The accepted papers will be included in the conference proceedings, and will be especially visible in the ACM Digital Library by being labeled with the appropriate “ACM Reproducibility Badge” .
The accepted contributions will have the opportunity to be hosted in the MMSys GitHub repository (source code) or in the ACM Digital Library or Zenodo (datasets).
The authors of accepted contributions will be invited to demonstrate their open source software package or a demo application using their dataset as part of the regular conference program.
Papers should be between two and six pages long (in PDF format) including references, prepared in the ACM style and written in English. Authors do not need to anonymize their submission due to the inherent difficulty of doing so for source code and datasets.

Regarding the submission of a dataset, the authors should make it available by providing a public URL for download. Regarding the submission of an open software package, the source code, dependencies on external libraries, and installation instructions must be available on a public web page or in a publicly accessible software repository. Additionally, the authors are requested to prepare a virtual machine image on which the software is pre-installed and ready for use or host the application online, unless the source code is self-contained and easy to compile without requiring administrator privileges or the installation of external dependencies. Please see here for instructions on how to prepare a virtual machine image. The URL to the dataset or software should be mentioned in the submitted paper. All datasets and source code must be licensed in such a manner that it can be legally and freely used, at the minimum in academic and research settings. Authors are encouraged to prepare as much documentation as possible to accompany the proposed dataset or source code, including examples of how it can be used by the public.

All submissions will be peer-reviewed by at least two members of the technical program committee of MMSys 2018. Criteria of selection include:

Datasets: The collection methodology and the value of the dataset as a resource for the multimedia research community.

Source Code: The broad applicability and potential impact, novelty, technical depth, and demo suitability of the proposed software.