AI Automatic Redaction | CaseGuard Studio Training Course
AI Automatic Detection in CaseGuard Studio
In this video, we’re going to explore AI automatic detection, which significantly simplifies the video redaction process. CaseGuard Studio leverages powerful machine learning technology, allowing you to perform redactions quickly and easily. With just a few clicks, you can automatically redact faces, license plates, vehicles, people, screens, checks, and signatures.
Adding Files and Starting Automatic Detection
- Add a File: Click the “Add files” button to select a video with many faces, which would take hours to redact manually.
- Activate Automatic Detection: Select the automatic detection button, represented by a smiley face on the left-hand side of your workspace.
Configuring Detection Settings
- Select Objects: In the automatic detection pop-up window, choose objects to detect—faces, screens, license plates, papers, people, vehicles, checks, and/or signatures. You can mix and match options to suit your needs. For this example, select faces and click next.
- Choose Shape and Effect: Stick with the ellipse shape for faces and blur for the effect. Adjust the effect level bar as needed.
- Redaction Types: Choose from standard (redacts detected faces), reverse (redacts everything but the face), or full-frame (redacts the entire frame when a face is detected).
Applying Redactions
- Select Range: Decide whether to apply redactions to the entire video or a selected range. For this example, choose “Whole File” and click next.
- Audio Options: You can mute, bleep, or resample the entire audio file.
- Classify Results: This option is automatically selected and helps identify each object detected.
- Confidence Level: Set at 20% by default; adjust if over-redacting occurs.
- Schedule Detection: Optionally schedule detection for later.
Click “detect” to start the process.
Reviewing and Adjusting Redactions
Once detection is complete:
- Playback: Click play to see how every face has been redacted automatically.
- Object Detection Panel: View each face identified and redacted. The number below each face indicates how many frames it was detected in.
Helpful Tools
- Go-to-Position Tool: Click this button next to any face to locate it in the video.
- Disable Redaction: Use the eye icon next to each object to disable redaction for specific objects. This is useful for excluding certain individuals from redaction.
You can also right-click on a redaction box and select “disable” to remove it.
Combining AI Detection with Manual Redaction
You can use AI detection alongside manual redaction. If any face is not detected:
- Pause the video and draw a redaction box manually.
- Use arrow keys to navigate frames.
- Once AI picks up your object again, right-click on your manual redaction box and select “Delete from next frames” to avoid duplication.
Exporting the Redacted Video
Export your video once redaction is complete. Notice how quickly you achieved results that would have taken hours manually. You can always make edits by accessing the original video in CaseGuard Studio.
Conclusion
That’s all for today! We’ve covered the three video redaction methods within CaseGuard Studio—manual redaction, object tracking, and automatic detection. In upcoming videos, we’ll explore audio and document redaction and showcase some of CaseGuard’s advanced features.Happy Redacting!