- Who should use WebPurify Image Filter?
- How does the WebPurify Image Filter Work?
- How does WebPurify compare to using other moderation services such as Mechanical Turks?
- What criteria do you use to moderate images?
- How long does is take for an image to be moderated?
- Who performs the moderation?
- How can I test the service during development without having to pay for moderations?
Q. Who should use this?
A. Anyone that is tired of moderating images uploaded to their site or application. Our service can be used to moderate any web hosted image, such as: avatars, profile pictures, contest entries, photo album pics, etc.
Q. How does the WebPurify Image Filter Work?
A.
1. You may use our standard criteria and begin submitting images right away or contact WebPurify to create custom moderation criteria.
2. You simply submit the URL to an image to our REST web service.
3. Your image will be moderated by our team of trained moderators.
4. The results of our moderations will be returned to a callback URL you provide.
Q.How does WebPurify compare to using other moderation services such as Mechanical Turks?
A.
1. We have many clients who arrived at our solution after being disappointed with M Turk. Error rates are quite high causing clients to have to run images more than once which impacts costs.
2. It is difficult to get consistent results with M Turk since guidelines for the unscreened freelance workers such as "nudity" or "sexually explicit" are interpreted differently and your only control over training is the ability to provide guidelines. An image reviewed for these criteria may be screened very differently by people of various ages and cultures around the world.
3. Managing workers requires a lot of overhead but with our solution we do all we can to be "fire-and-forget".
4. The biggest concern we have found with clients is the exposure of their image to an unscreened freelance worker who can easily capture it and disseminate it with no liability. Privacy in this space is extremely important.
Q.What criteria do you use to moderate images?
A. By default our service flags images that contain “Adult Content”
We define this as any image that may :
• contain nudity or partial nudity?
• portray hate or hate crimes?
• contain violence or show weapons?
• contain offensive gestures or language?
• contain drugs, drug paraphernalia, or drug use?
• appear blank or broken?
However we also provide custom solutions, please email support@webpurify.com
Q. How long does is take for an image to be moderated?
A.Ranges are usually from 1 minute to 5 minutes from the time you submit the image to our service. During times of high volume can go as high as 10 minutes.
Q. Who performs the moderation?
A. Our very qualified workforce that is prepared to handle your requests 24/7 365 days a year.
Q. How can I test the service during development without having to pay for moderations?
A.We provide two methods for testing. One for testing integration and another for testing accuracy.
Testing Accuracy:
We provide 50 free moderations for new accounts.
Testing Integration:
We have a sandbox method: webpurify.sandbox.imgcheck which can be used to check that your integration is working correctly.
All images submitted to our service using this methods will bypass our virtual workforce and be moderated automatically (randomly). The service will hit your callback URL to complete the process. It is important to know that this call may result in false positives as it is bypassing our moderators.
Standard Image Moderation Criteria
Does the image:- contain nudity or partial nudity?
- portray hate or hate crimes?
- contain violence or show weapons?
- contain offensive gestures or language?
- contain drugs, drug paraphernalia, or drug use?, or
- appear blank or broken?

