Scenario

You work for a marketing company. Your company receives customers’ information and photos, curates them, and creates social media profiles for them.

However, your clients complain that they are not getting enough hits. After a study, the data science team attributes the lack of hits to unprofessional-looking photos.

You are tasked to create a system that identifies if a photo is professional-looking enough before it is uploaded to social media.

Requirements

You want to build something that evaluates if a photo is professional-looking. But what does it mean to be professional-looking? You decide that professional-looking photos require the subject to be smiling and have their eyes open.

How can you identify these characteristics in a photo? It is not an easy task to program logic that identifies specific characteristics in photos, especially when photos can come in so many file formats. An ML model has the best odds of yielding good results.

Because...