Why is face detect not working?

Face Detect uses AI to find a face in an image and position it in the center of the container. There are common instances where the result might not be within your expectation.

Inaccessible image URL

The most common issue is that your input URL might not be accessible at the time you sent your request, blocking Bannerbear from accessing the file, or the host is slow to respond. Try to check your host settings or try a different host for your input.

Image Format

Bannerbear supports a wide range of image file formats, but we recommend using JPG or PNG files.

No faces detected

Your input might not have a well defined face or does not have any face. Ensure that your input has a distinguishable face subject. If you're trying to focus on objects, Smart Crop is more suited for it.

Multiple faces detected

Your input might have multiple faces being detected. In such cases, Face Detect will try to set a focal point between the faces and focus on that. Try a different input with a single subject or set a custom Anchor.

Excessive Zoom

Face Detect is set to apply the least amount of zoom necessary to center the face and fill the container. The position and size of the face within the image input may affect the zoom level (eg. face is located at the edge or corner).

In addition, the orientation of your template and image input will have different results (eg. square image with a large face to an Open Graph sized template).

Ensuring that your input has enough buffer around the face and matching the orientation of the template and input can have better results.