|
Posted on behalf of Bob Planqué and Willem-Pier Vellinga. If you have questions, please reach out: This could be a game-changer (great work, you guys!), and I encourage developers working on annotation tools to ensure compatibility with
XC. Thanks!
******************************************************************************************************************************************
Hi, everyone!
It's been a long-standing plan: annotate individual sounds on Xeno-canto ("XC"). We're making it happen now.
Recordings on XC are typically attributed to a single species, the "foreground" species. Additional information ("metadata") such as sound type, sex, and more can be provided. However, many recordings on XC contain multiple vocalizations, often from multiple species.
The single “label” for the recording (ID+type+sex, for example) isn’t sufficient in these cases. Additionally, this label doesn't clearly specify when and where the sound occurs, or its frequency. It is a "weak" label.
The lack of precise temporal and frequency information, along with missing metadata for sounds within each recording, means that much potentially interesting detail in the collection is currently hidden: hidden sounds, hidden behavior, hidden diversity. It
is also problematic for all kinds of analysis, especially when multiple species are audible within a recording. In fact, the lack of detailed information on the exact timing and frequency of vocalizing species in recordings is currently the main obstacle to
developing high-throughput automated analysis (e.g., modeling call sequences, automated recognition in large soundscapes datasets, and so on).
However, that situation can now change because XC now allows uploading annotations (bounding boxes) for individual sounds in recordings. We have added features to upload annotations to XC, download annotations from XC (also via the API), search for annotations
on XC, and view annotations in a new recording player.
Take a look at these articles for more details:
We are currently in a testing phase, and the feature to import annotations into XC is available to a limited group of people involved in its development. If you'd like to assist with testing and start annotating recordings on XC, contact us. Once we confirm
everything runs smoothly, we will open the annotation upload to everyone.
When we launched Xeno-canto, we announced it on NEOORN and closed with a sentence that still feels appropriate: "We hope that over time it will turn out to be a useful and enjoyable resource, and we also hope that it will generate a bit of honest web community
effort. You're all invited to have a look around and participate!"
Bob Planqué and Willem-Pier Vellinga
--
At the Yang Center, we work flexibly, and while it suits me to email now, I don’t anticipate a response outside your regular work hours. Thanks!
--
Dr. Holger Klinck
John W. Fitzpatrick Director
K. Lisa Yang Center for Conservation Bioacoustics
Cornell Lab of Ornithology, Cornell University
Faculty Fellow
Atkinson Center for a Sustainable Future, Cornell University
Courtesy Professor
Marine Mammal Institute
Department of Fisheries, Wildlife, and Conservation Sciences, Oregon State University
Mailing address:
159 Sapsucker Woods Road
Ithaca, NY 14850, USA
Tel: +1.607.254.6250
Fax: +1.607.254.2460
Email: m("cornell.edu","Holger.Klinck");">
Web: https://bioacoustics.cornell.edu
How I say my name
Pronouns in use: He/Him/His
This email and any files transmitted with it may contain confidential information. If you believe you have received this email or any of its contents in error, please notify me immediately by return
email and destroy this email. Do not use, disseminate, forward, print or copy any contents of an email received in error.
|