Question

A social marketplace app allows users to upload product photos. The product team needs an automated way to flag images containing adult, violent, or racy content before the images are published, with minimal development effort and no model training. The volume is unpredictable and spiky. Which Google Cloud approach best meets these requirements?

Accepted Answer

Use the Cloud Vision API SafeSearch Detection feature to score each uploaded image for adult, violence, and racy likelihood.

Answer

Train an AutoML Vision image classification model on a labeled dataset of safe and unsafe images and deploy it to a Vertex AI endpoint.

Answer

Use Document AI to parse the uploaded images and extract text for policy violation review.

Answer

Build a custom TensorFlow CNN, train it on Vertex AI, and serve predictions from a Compute Engine instance group.