FAQ
I’m new to this project. Where do I start?
We’re glad you’re here! To learn more about this project, please visit the research page. When you’re ready to participate, go to the main project page and pick a workflow. The Tutorial window will appear. Read it very carefully, then start classifying! Feel free to contact us on Talk if you have any questions.
What if I’m not sure about my response?
Don’t worry! We’ll combine your answer with the answers of other volunteers, so just try your best to make the correct classification. If you get stuck, there are three resources on the classification page that might help:
- The Field Guide has helpful information and example vocalizations for each of the species you’ll listen for throughout the project. We did our best to assemble a comprehensive Field Guide, but occasionally you may be unable to verify a recording using only our example sounds. If this happens, scroll to the bottom of a Field Guide entry and visit the linked resources under the “For More” heading. These resources will connect you to even more example vocalizations to help you confidently make a classification. You can access the Field Guide as many times as you need to, anytime throughout the project.
- The Tutorial has information about spectrograms and the instructions for your task. You can access the tutorial at any time by clicking the “Tutorial” tab on the classification page.
- The “Need help with this task?” button below the answer options on the classification page has some basic reminders about your task and the resources available to you.
If you make a mistake, that’s okay! While you should try to do your best, mistakes happen. Keep going—we promise, this will get easier with time.
I can hear multiple bird species in the recordings, but the questions only ask me about one species at a time.
For this project, we’re asking users to verify sounds from one species at a time. This means that if you hear birds other than the target species, you can disregard them. You only need to focus on the target species. That said, if you’ve confidently identified a notable non-target species and want to share your findings with the research team, please feel free to do so with the “Done & Talk” button on the classification page.
I’ve been marking “yes” on a lot of the recordings. Am I missing something?
The AI model that is classifying these sounds, BirdNET, is better at some birds than others. For some species, its detections might be mostly correct, while there might be more false positives for other birds. So, if you find that you’re marking “yes” on most recordings for certain species, you’re probably right on track! What’s important is that you still check each recording carefully for misclassifications, even for species that the model tends to identify correctly.
If computer models can identify bird species from their sounds, why do you still need our help?
Scientists have recently developed artificial intelligence software that will estimate which bird species are vocalizing in recordings. Tools like this are useful and exciting, but they can make mistakes. By verifying detections, you’re helping us find mistakes so we can correctly interpret model output.
I hear something interesting in the recording that I want to bring attention to.
Want to share a sound? We’re listening! After making your classification, please click the “Done and Talk” button and “Add a note about this subject”.
I want to learn more about birds, pyric herbivory, or AI sound analysis.
To find information on bird identification, importance, and ecology, check out the Cornell Lab of Ornithology’s All About Birds website.
Learn more about The Prairie Project at our website.
Our project uses the open-source AI sound analysis tool BirdNET, a collaboration between the Cornell Lab of Ornithology and Chemnitz University of Technology. You can try BirdNET yourself with the BirdNET mobile app, or try a similar feature using the Cornell Lab of Ornithology’s Merlin Bird ID app.
I have specific questions about this project.
You may contact the research team on our project Talk page. We’d be happy to answer your questions.
I have general questions about technology, Zooniverse, or my account.
If you think your technical issue is specific to this Zooniverse project, get in touch with us on our Troubleshooting discussion board on Talk. Otherwise, please refer to Zooniverse’s FAQ or knowledge base, or contact the Zooniverse team if needed.
What’s going to happen to the sounds I classify?
By verifying the results of the AI model, you’re telling us what species it tends to identify correctly, and where it makes mistakes. This information will help us accurately interpret model results on a long-term basis. Some of the sounds you classify will become part of a database of bird sounds to help scientists preserve, share, and study the bird sounds of Texas grasslands.
What do the tick marks on the y-axis of the spectrogram indicate?
The y-axis of a spectrogram indicates frequency, or the pitch of the sound. This is measured in kilohertz (kHz). The two tick marks on the y-axis are at five and ten kHz. For more on spectrograms, please read the project’s Tutorial, available on the classification page.