What’s the Real Gender of an Author? and Facial Gender Analysis

There are a few fun tools available on the web these days that might be of particular interest to those of us in the transgender world.  The tools themselves are designed to be more fun than informative and were probably not designed specifically for transgender people at all.


The first one I’ll talk about today is Genderanalyser.  If my memory serves me correctly, the media story about this mentioned that it was written at Harvard or one of the big US schools.  It’s purpose in life is to tell if the text at the user-provided URL was written by a male or a female.  It provides a likelihood percentage and asks us to tell it if it’s correct.

“There are those who wish there were no differences between men and women. In the 1970’s at the University of California, Berkeley, the buzzword among young women was “mandatory unisex,” which meant that it was politically incorrect even to mention sex difference.” ~ Louann Brizendine, M.D.

I provided it with the url for my About Me page and it believes there is a 72% likelihood that page was written by a female.  When I tried my blog post Could I Blog Professionally? it replied “We guess http://seleenak.com/could-i-blog-professionally/ is written by a man (56%), however it’s quite gender neutral.” 

I was a little surprised that my blog post New Website Look: The Reason For the Change returned 59% female.  I would have suspected that the frustration I was feeling when I wrote that would have come out as male characteristics but they really didn’t.

You can find Genderanalyser here.  Give it a try!


The second tool I’ll mention is Pictriev,  the Facial Search Engine.  Upload an imagine or provide an image URL and Pictriev will try to tell you if the face is male or female. It will also try to guess the age and provide a short list of famous people with similar faces.

I provide this image. It found my face buried in all the hair and scored it 22% masculine, 78% feminine and estimated my age at 26 – I’m more than double that age. Lookalikes include Paula Abdul, Joan Collins, Angela Lansbury and Jackie Collins.
Then I uploaded this one. It scored 61% masculine, 39% feminine and age 40. Lookalikes include Oprah Winfrey, Naomi Campbell, Young Jeezy and Lenny Kravitz.
Lastly, I tried this one. It scored 68% masculine, 32% feminine and age 45. Lookalikes are Susannah York, Sasha Cohen and Bianca Jagger. Interestingly, it believes the face is mostly male yet found only female lookalikes.













I also tried uploading a couple of photos of genetic females and, here are the results:

Marilyn Monroe is 0% masculine, 100% feminine and age 29. She died at age 36 so even the age is entirely possible. Her lookalikes were Marilyn Monroe (ha!), Barbara Lawrence and Mamie Van Doren.
Audrey Hepburn, who some believe had the perfect face, didn’t score quite as well as Marilyn. This photo as rated 6% masculine, 94% feminine and age 26. Her lookalikes were Audrey Hepburn (yay!), Sophia Loren and Jessica Marais.













So there you go.  Although I don’t know how the algorythm works, I think, if someone was so inclined, they could analyze their own results and make adjustments to their look to move the percentage in the direction of their choice.  A good self-improvement tool.

Yet when I discussed this online with tgirl friends their photos seemed to be scoring very high .. much higher than mine.  When they suggested I keep trying different photos until I got a high feminine rating I realized what they had been doing.

Instead of considering PicTriv fun or informative, they were using it flatter themselves.

We’re such an insecure, delusional group, aren’t we?  But maybe that’s what makes us sweet!

Leave a Reply