Tag Archives: Hidden Talent Show

Another super-recognizer test, just a wee little one

Maybe this is the super-recognizer test which lots of my recent blog readers were looking for. You don’t need to be logged on to Facebook to do it. It appears to be the creation of the UK face recognition researcher Dr Ashok Jansari and his team at the University of East London and it is presented by the UK TV show Hidden Talent. I had a go at this test and got a score in the normal, not super-recognizer range. It is a tricky test, requiring the person doing the test to identify faces seen for only seconds in two quite different emotional and visual contexts. The test is designed so that non-face elements of a person’s appearance such as hair cannot be used to identify, thus it is a true test of face recognition and can’t be cheated by using memory for other elements, and this test also it isn’t just a test of photograph recognition, which is a criticism that can be made of some other tests that only use one photo of each face in the test, with photos often including hair and other background elements. One good thing about the test is that it includes faces of both sexes, which possibly makes it a more realistic measure, while some tests of face memory used by university researchers include only male faces, including the respected Cambridge Face Memory Test. People taking this test are required to memorise faces shown from a particular angle and displaying a particular emotional expression, and are later required to identify some of these memorized faces shown in a different angle and/or a different emotional expression. This might seem like a realistic way to test face recognition but I doubt that it is, because in real-life situations even if we only meet a person for minutes or seconds we usually get the opportunity to memorize the appearance of a face across some kind of range of angles and expressions. This is not the same situation as being required to recognize a face across a range of angles and expressions. In this test I think the phase of face memorization is limited compared to real-life situations of face memorization. If the difference between a natural super-recognizer and a normal recognizer lies in the richness of the encoding of the memories of faces, then this test might not be fit to measure this. I believe the fairly artificial limtation of the memorization phase is one of this test’s flaws, and in this respect it reminds me a lot of the second test of face recognition which I was given to do when I volunteered as a research subject in an Australian university in 2011. I don’t know the official name of that test and I was never informed of my score in that test. In that test I was required to memorize Chinese-looking male faces in profile and identify them displayed in a full-face angle, and it just didn’t feel like face recognition. I know that any success that I had in that test was probably due to employing conscious strategies for face matching (such as making conscious note of facial features and matching skin colours), which most certainly isn’t natural face recognition (which is a completely automatic and unconscious process, rather like synaesthesia).

One could also definitely criticise this test for not being large or long enough, and thus more likely to give results biased by chance. There are only 11 faces presented for memorization, to be picked out of a set of 15 faces presented in the second part of the test. Compare this with the short version of the Cambridge Face Memory Test with a maximum possible score of 72, and it’s clear that this test is not a lot more than a bit of fun. One definite problem with this test is that I found that I could get a score in the normal range using a very simple strategy without even looking at the faces. I’m wondering how anyone could get a score in the low range, and this test appears to have no value for identifying prosopagnosia. A criticism that could be levelled at all tests of face recognition or face memory is that they don’t reflect real life face recognition situations. When we meet people, even if it is just for a few seconds, we usually see a moving, speaking image, not a still image, and in that movement we see not just a face from a range of angles but also the accompanying body language, probably a range of different emotional expressions, and also the very individual ways in which a person moves their face and body. When you meet a person you see the life and the personality in their face and body, not just a static piece of meat, and that is more memorable than a still image of a face. I’m wondering why face recognition researchers haven’t come up with a test that uses video clips rather than still photographs. It seems like an obvious way to make a face recognition test more like a test of what people need to be able to do in real life.

Super recogniser. https://apps.facebook.com/hiddentalentshow/fb/tests/recognizer

What happened yesterday?

Wednesdays are usually good days for reader stats but yesterday there was an extraordinary spike in stats for this blog. It was a large amount of traffic from the UK, and appears to be a media-driven upswing in interest in testing relevant to super-recognizers or superrecognition. There is definitely an international interest in an online test specifically designed to sort super-recognisers from normal people which is culturally-neutral and substantial enough to give meaningful results. It’s a pity such a thing does not appear to be freely available.

Hidden Talent Series 1 Episode 5 May 22 2012 Channel 4 UK  http://www.channel4.com/programmes/hidden-talent/4od#3340727

Super-recognizer to appear on Channel 4’s The Hidden Talent Show? Documentary about superrecognizers?

From Facebook three weeks ago: “For a Channel 4 programme called Hidden Talent Richard Bacon interviewed UEL Lecturer Ashok about his study into Super Face Recognition and it’s polar opposite Face Blindness”

From Twitter February 24th 2012: “Dr Jansari currently being interviewed by Richard Bacon for Channel 4 documentary about super-recognizers.”




View the show here for a limted time, subject to geoblocking:  Hidden Talent Series 1 Episode 5 May 22 2012 Channel 4 UK  http://www.channel4.com/programmes/hidden-talent/4od#3340727

From the written description of this episode, it appears that Dr Ashok Jansari administered the Before They Were Famous Test and also probably the Cambridge Face Memory Test to a crowd to find super-recognizers, a screening process modelled on the 2009 paper by Russell, Duchaine and Nakayama which launched the concept of the super-recognizer. This screening process identified three super-recognizers: Richard (male name), Charlie (gender not clear) and Higo (male). They are then put through a live test of face memory.