Using background image(No object image) for validation data while Azure AutoML Object Dectection Train

### Describe your suggestion

Hi
I want to check azure automl object dectection model's False Positve performance whie model training.
so i added background image(No object image) to validation data
in confusion metric of model's metric page, but there is no predict result for background image.
(sum of confusion metric value is same with number of object image(except no object image)

Please let me know how to add background image(No object image) for perfomance metric.

### Additional details

_No response_

Start by reproducing Azure AutoML object detection training with background images in the validation data and inspect the confusion metric page shown in the issue. Determine whether background images are expected to produce a validation result and identify the repository entry point, example, or documentation that governs this behavior. Done means the supported way to measure false positives with no-object images is documented or implemented.

Written by the indexing model from the issue text.