Using image analysis tools to avoid scams
Todays #1 job search problem
One of the biggest problems for online job searchers this past year has been that they have run into more scams than ever. I forget the exact percentage that the BBB was quoted as giving for their scam-to-job ratio, but I do recall that it was extraordinarily high. I believe it was well over 90% of potential jobs that job seekers found online turned out to be one sort of scam or another.
Of course there are the usual scams like envelope stuffing jobs, advance fee fraud, and fake companies posting as mystery shopping opportunities.
But some of the newer scams are a bit harder to pin down and take advantage of the worldwide nature of the internet and the belief that many people new to the internet have that the online world somehow operates significantly differently than the real world.
So the scammers come up with realistic sounding jobs like rebate processing, web processing, or other types of "positions" that sound like real work, but actually mean little or nothing at all. Or sometimes just aren't at all what they sound like they should be.
Using reverse-image search engines to ferret out fraud
One quick step you can take to ferret out fraud on a lot of sites that seem too good to be true is to use a reverse-image search engine to determine if the testimonials on the site are potentially fake. Of course, you can ferret out a lot of fraudulent job opportunity sites by the mere fact that they are charging a fee for a job, but many sites have become more sneaky about how they charge people and many people are getting charged before they even realize it.
Besides, this technique can be useful to detect potential fraudulent marketing of almost any type of product, service, or opportunity so it's worth learning for that sake a lone.
What you will want to do is use a reverse-image search engine to determine if testimonial pictures used on a site are potentially fake. Often if they are fake, you will be able to find them on a stock photo site. You can see an example of where I did this with the site DataMoneyOnline here.
By using the reverse-image search engine Tineye I was able to quickly determine that almost all of the testimonials listed on that site were in fact stock images.
Comparing two images
Sometimes it can be valuable to check for the difference between two similar images to see if there has been image manipulation of the original. Of course this depends on determining which image was the original in the first place.
A useful tool for such analysis is the tool phash.org which describes its strong point of perceptual hashing as:
"Perceptual hashes must be robust enough to take into account transformations or "attacks" on a given input and yet be flexible enough to distinguish between dissimilar files. Such attacks can include rotation, skew, contrast adjustment and different compression/formats. All of these challenges make perceptual hashing an interesting field of study and at the forefront of computer science research."
For our purposes, we just want to know if a file has been changed or altered in any way.
Search images by faces
One type of tool I've had less success with are face search engines which allow you to search by faces. Maybe I'm asking for too much or maybe I don't understand exactly what face search engines are trying to accomplish.
I've tried to use them when I've been unsuccessful in finding a duplicate image using Tineye and to this day have never been successful.
The one I've tested the most is Picollator.com which is a face search engine developed by the company Recogmission, LLC.
No comments yet.