Troubleshooting Wyze Facial Recognition Issues
Wyze's Facial Recognition feature, part of the Cam Plus subscription, adds a powerful layer of intelligence to your security camera, allowing it to distinguish between friends, family, and strangers. When it works, it's brilliant. When it doesn't, it can be confusing. This guide will help you resolve common problems with Wyze Facial Recognition.
## Prerequisite: Confirm Your Subscription and Settings
Before you start troubleshooting, you need to ensure the feature is properly set up and enabled. Facial Recognition is not a free feature; it requires an active Cam Plus or Cam Plus Pro subscription.
- Check Your Subscription: Open the Wyze app and navigate to the Account tab. Go to Services and verify that you have an active Cam Plus subscription assigned to the camera in question.
- Enable the Feature:
- From the camera's live stream, tap the Settings gear icon.
- Go to Event Recording.
- Tap on Smart Detections.
- Ensure that Person Detection is enabled. You will not get face detections if person detection is off.
- Enable Facial Recognition:
- Go back to the Account tab.
- Tap Services > Cam Plus.
- Find your camera in the list and ensure the Facial Recognition toggle is switched on.
## Step 1: Building Your Face Collection
The system cannot recognise faces it hasn't learned. You need to actively teach it who is who.
- Capture Clear Images: For the initial setup, have the person stand in a well-lit area and look directly at the camera. Make sure their face is not obscured by hats or shadows.
- Tag Faces from Events:
- Go to the Events tab in the Wyze app.
- Find a video clip where a person's face is clearly visible.
- Below the video, you should see a thumbnail with "Is this...?" and an option to "Name this face".
- Tap this, and either assign it to an existing person or create a new profile for them.
- Tag Multiple Angles: The more you tag, the smarter it gets. Tag photos of the person from different angles, with and without glasses, and in different lighting conditions to build a robust profile.
## Step 2: Optimise Camera Placement and Lighting
The performance of facial recognition is highly dependent on the quality of the image the camera captures.
- Lighting is Key:
- Avoid strong backlighting where a person is standing in front of a bright window or doorway. This will turn their face into a silhouette.
- Ensure the area is well-lit. In low-light conditions, the camera may switch to black and white night vision, which is less effective for facial recognition.
- Camera Angle and Distance:
- Position the camera at roughly head height. A camera pointing down from a high ceiling will mostly see the top of people's heads.
- The person should be within about 15 feet of the camera for optimal recognition. If they are too far away, the facial details will be too blurry for the AI to analyse.
- Clean the Lens: A smudged or dirty camera lens can soften the image and dramatically reduce accuracy. Clean it regularly with a microfibre cloth.
## Step 3: Correcting Mistakes to Improve the AI
When the system gets it wrong, you need to correct it. This is a crucial part of the learning process.
- Fix Misidentifications: If the camera tags your friend Bob as your friend Steve, go to that event video. There will be an option to report an incorrect face. You can then re-assign the image to the correct person's profile.
- Merge Duplicate Profiles: Sometimes, you might accidentally create two profiles for the same person (e.g., "John" and "John Smith"). In the Facial Recognition settings, you should have an option to merge these duplicate profiles into a single one, consolidating all the learned images.
By actively managing your face library, optimising your camera's environment, and consistently correcting errors, you can significantly improve the accuracy and reliability of the Wyze Facial Recognition feature.