To Do

Train/Test dataset

• Collate training clips and place in one location where all authors have access

Evaluation dataset

• Decide scope of evaluation dataset and identify periods of interest
• Identify any additional annotation that needs to be done within the evaluation dataset

Administrative

• Data sharing agreement from each contributor
• Publication plan
• Provide links to other public datasets that may be of interest to participants (for training)

---

Data Sharing Agreement

Principles

Open Access - We believe that conservation is a team effort and that the field can be moved forward by building ecologically representative data and sharing those data both within and outside of the field.

Acknowledgment- Building, maintaining, and annotating acoustic datasets is labor-intensive and costly. As there is no financial compensation for contributing data to this project, all contributors should be acknowledged in any and all products arising from this work. More directly, it is inappropriate to use these data without proper attribution.

Collaboration- Beyond acknowledgment, the maintainers of the included datasets are experts in the field and the most knowledgeable sources regarding the specifics of and limitations within their data. Data contributors should be invited to participate an any studies utilizing these data.