I'm looking to create a binary classifier for images.
Basically, I want to detect if an image contains a person or not. I've trained the model with photos of that person, and photos that are NOT that person. I generally get around 85% even when I limit the images to faces. Firstly, am I using the trainer correctly, to train a binary classifier as the person or not, and secondly, how can I improve the recall/precision?
