I've been trying out Google's Cloud Vision API (https://cloud.google.com/vision) as a means of detecting NSFW content in user-submitted images but am finding it surprisingly unreliable. A large proportion of requests appear to just randomly hang before timing out with no error. I'm looking for recommendations for an alternative solution which can flag images containing pornography, gore, violence, etc. All the bad... - Source: Hacker News / 9 months ago

I wonder if we could use something like https://cloud.google.com/vision to employ AI image classification on these sat images? Source: 11 months ago

There are many. Google Lens can classify images. It can even identify species of plants and animals. If by "toolset" you mean you want an API to write your own applications, you can use the Google Vision API. Source: 12 months ago

Could you use Vision AI for this maybe? Source: about 1 year ago

The violence will get it not recommended, use https://cloud.google.com/vision/ to see what I'm talking about. It comes up as very racy and possibly adult. Source: about 1 year ago

Another annotation tool that integrates prediction and training within the application is supervisely supervisely.com., unfortunately it's pretty expensive unless you are satisfied with the community version. I saw that they have an integration for owl-vit, which might be helpful for annotation of animals. https://ecosystem.supervisely.com/apps/serve-owl-vit. Source: about 1 year ago

Hello world. This tutorial is a gentle introduction to building modern text recognition system using deep learning in 15 minutes. It will teach you the main ideas of how to use Keras and Supervisely for this problem. This guide is for anyone who is interested in using Deep Learning for text recognition in images but has no idea where to start. - Source: dev.to / over 1 year ago

If they were videos, I would have suggested trying supervise.ly as it has a very good tracking functionality. Source: almost 2 years ago

Hi, I'm exactly in the same boat like you are. I looked around for a while and the better solutions I found was supervise.ly and CVAT for video annotation. The pricetag on supervisely is pretty high, so I analyzed CVAT for a couple days and was positively surprised. Source: almost 2 years ago

Under the WPI Photo Ambum section of the page for FRC field photos (https://www.firstinspires.org/robotics/frc/playing-field#WPIPhotos), they have a section of machine learning imagery. However, this link goes to supervise.ly, the website they use for machine learning. I created an account to attempt to download the images, however, whenever I try to 'clone' the project, it stalls at 0% and gives me an error... Source: almost 2 years ago