Help

A ten-step guided tour of how Sparrow Studio turns camera and audio recordings into reviewed wildlife data, from the first upload to an exported result.

You can step through this tour now, skip it, or reopen it any time from Help.

A Project is the container for one study. Everything below lives inside it.

• Project - the study itself, holding everything below
• Locations + deployments - where your cameras are placed
• Ingestion - how images, video, and audio reach Sparrow Studio
• AI detections - the model's predicted species for each capture
• Review - where you confirm or correct those predictions
• Explore + export - filtering, charts, and reports for what you've reviewed

Create the project first, then configure its connections, devices, processing models, API keys, and tracking integrations from Project settings.

The Dashboard lists the projects you're authorized to see. This step is a read-only look at that list - nothing is created for you.

You'll see a project here if you created it or you've been added to it as a project member; administrators can see every active project.

Field gear is organized as a chain, from a place in the field down to a single captured observation.

• Location - a place in the field
• Deployment - one camera active at that location for a date range
• Camera - the physical device
• Observation - an image, video, or audio clip that camera captured

How media reaches Sparrow Studio depends on how your project is set up: a real-time email inbox, FTPS, a live RTMPS stream where that's configured, a Sparrow Edge or Water Sparrow device and its API, or a manual upload of image, video, or audio files. This step is read-only - nothing is sent.

Whichever channel is configured for your project, incoming media is tracked before it moves on to AI processing.

Once media arrives, an AI model attaches a predicted species label and a confidence score to each detection.

These are predictions, not final scientific answers - that's what the next step is for.

In the workspace, every detection shows the model's predicted label next to the space for your decision.

Once you've confirmed or corrected the detections on an image, mark that image Reviewed; Unreviewed images stay easy to find with the workspace filter.

Filter your results by date, location, deployment, species, or review status, and see the counts summarized as charts.

Filters apply to your whole media set - reviewed and unreviewed alike - unless you narrow it down by review status yourself. The same filters you use to explore results carry through to export.

Add a reviewer note to any detection, and @-mention a teammate by username to flag it for them.

A mention creates an in-app notification on the Messages page - mentions do not send email or push notifications.

Project alerts are different: an alert can have its own configured recipient emails, and may send email to them when a detection matches, if alert processing and email delivery are set up and running for your deployment. Check with your team rather than assuming either way.

Export reviewed detections as CSV, or generate a biodiversity report, whenever you're ready to share results outside Sparrow Studio.

You can reopen this tour any time from Help.