It detects adult content, but does it detect illegal content as well?

Question

Hi,

I published an image sharing Web application that converts 360 Images "ready-to-share" and embeds them for the Web. That works very well, but I need to moderate the uploaded contents. I do not want anyone to abuse my platform for unwanted or illegal content publication.
I already have implemented that Cognitive Services detect adult content so I get a notification and can block those accounts if necessary. But what about illegal images? Is it possible to detect them too?

With illegal and unwanted content, I think about pictures that show

• Illegal Symbols
• Children pornographic content
• Racism
• Brutal and violent content

Can I detect this anyhow?

Best Regards

Answer 1

@Marco Griep Thanks for the question. Currently the following is supported, Please feel free to raise an uservoice for other illegal and unwanted pictures.
Computer Vision has the capabilities to detect adult/racy/gory images in images, but we haven't yet provide any workflows around those capabilities. The underlying models for Adult/Racy are the same between the two services(computer vision and content moderator APi).

The Content Moderator Api doesn’t offer the Gory classification as yet.

I see both can detect adult and racy content in images, but the Content Moderator also supports video and text processing, as well as workflows.

you can combine CM(content moderator) with your custom developed APIs (mainly in Azure ML) for custom detection (image quality and content etc.,) since CM workflow provides APIs, when connecting to 3rd party API as you can see only these connectors, you can work only through API/SDK if anything else needed.