Is it possible to train the OOB ObjectDetection

Hi,

Is it possible to Train the ObjectDetection Out of the Box model to be able to identify certain objects in a picture?
If yes, then what is the minimum number of images we have to use in the Training and Evaluation Pipelines?

Thanks in Advance

@hgaber105

Minimum of 10 images is recommended to train and evaluation can have as low as 2 as well

Cheers

1 Like