Recently in Face Recognition Category

This article was a study that was completed at Iowa State University.  It highlights qualities of identification testimony, feedback to eyewitnesses, On-Line vs. post-computed judgments, and securing false identifications.

This article relates to everything we have previously talked about in class.  I really recommend that you take the time to read through it and post any comments or questions about the study that come to your mind. 

http://www.psychology.iastate.edu/FACULTY/gwells/Wells_articles_pdf/Good,_You_Indetified_the_Suspect.pdf

For my book report for this course I am reading Mind Hunter by John Douglas and Mark Olshaker.  In the book, Douglas describes his interviews with criminals in the prison system in order to develop profiles for serial killers as part of a huge research project.  This was part of an effort to develop a Psychological Profiling Program, to make it easier for investigators, police, and the FBI to catch criminals faster and easier, by being able to "get inside their heads" and track their next moves. This got me interested in criminal and psychological profiling so I decided to google it.  I came across digital physiognomy software and free downloads.

 

 

Digital Physiognomy was developed as entertaining software, allowing people to create a portrait of a person and then provide a description of the person's character, traits, preferences, likes and dislikes, etc. You can choose different facial features and put them together, similar to the composite software that we can borrow from Dr. Maclin for our projects and what police use to make sketches of suspects. 

Testing out digital physiognomy or composite software gives us a real life account of creating a "picture" of a suspect.  I checked out the composite software from Dr. Maclin and found it difficult to do.  Difficult for the "witness" to come up with a description and difficult for me to choose accurate features from the extensive list.  Of course composites are biased to either the witness's perception or memory and the composite creator's training and expertise; however, both are useful in coming up with a tangible idea of the suspect. 

A free demo version is available at http://www.uniphiz.com/.  I did not have a chance to download it or see if it works, but I encourage people to check it out and see what it is all about!

 

 

This link will take you to a story featured in FRONTLINE, as a result of a mistaken identification from an eyewitness.  In the story, Jennifer Thompson was trying to identify her rapist, but because of faulty lineups, she chose an innocent man instead of her actual rapist. It really is an interesting story because they give you the lineup with the mistaken rapist and they also show you the real rapist. It also gives the composite, and how the two relate to it. 

You should check out the website and click around on some of the links about the case and what Thompson saw in the lineup. 

This link will take you to an article that went along with this story and goes along with what we've been covering in class about constructing lineups. 

The article is written by Gary Wells, a psychology professor at Iowa State University.  He, like Dr. Maclin, believes that there are faults in our legal system because of the methods of eyewitness identification.

As this case and prior data and findings suggest, several people are wrongfully convicted based on wrong eyewitness identification and faulty lineups.  The two play off each other as well.  For example, say a witness believes they can identify their attacker, however the police constructed a lineup where one suspect completely stands out from the rest, therefore the witness is more compelled to pick the person who stands out more, not realizing that this is actually not the man who attacked her. 

The witness's perception could also bias their pick out of a lineup.  If a guy looks meaner than the rest, or has more tatoos, the witness may associate this person with being more aggressive and therefore, more likely to be the offender.  As this area of research continues to grow and collect more results, better methods of obtaining eyewitness information and constucting lineups are being developed in hopes of improving of our justice system.  The growth of DNA testing and its use as evidence is also becoming more widespread and accurate to put the right people behind bars. 

http://www.keysso.net/commrelations/crimepreprograms/suspect_id.pdf

Monroe County Sheriff's Office in the Florida Keys has a big population with few Sheriff's to cover all of the people. On their web page it states, "the Sheriff's Office employs 578 people. 187 of those are road patrol officers and detectives, 144 are Corrections Officers, 182 serve as support 65 people are employed by the Sheriff's Office-managed HIDTA Group (High Intensity Drug Trafficking Area)."  They have this many people for as many as 150,000 people in their population and that does not count the tourists that can reach as many as 2,000,000 people.  With so many people in their population, and a high turn-over rate the department needs something to help catch suspects.

To help identify a suspect of a crime, the Monroe County Sheriff's Office has a document that allows the witness to fill in details of the perpetrator they saw. The document has the witness fill out the physical appearance of the suspect (with a drawing of a blank suspect to the right), the clothing they were wearing, what weapon was used and there is a drawing of weapons at the bottom of the drawing of the person, and anything about the vehicle used in the crime (if one was used).  It is a very detailed outline asking anything from the scar marks, speech, or physical defects of the perpetrator,  to the license plate number of the vehicle used. 

I believe this would really help local law enforcement agencies around Iowa to catch the correct people, instead of mistakenly persecuting innocent people.  This document can come in very handy if someone gets victimized and needs to write down anything they can remember of the perpetrator, the weapon, or the vehicle used in the crime.  I believe having this handy will allow anyone to keep their memory less contaminated if they write all of these facts down right away before talking to anyone.

During the past two weeks of class, we've been discussing both the power of eyewitness testimonies, but also how unreliable memories can be. People do not realize how difficult it is to accurately identify someone even when a witness may be paying close attention to their surroundings.

This activity allows you to play the part of an eyewitness and accurately "describe" the person. Although incredibly over simplified, it illustrates that it's not as easy as it may seem. I did the exercise a few times, varying the amount of time I looked at the top picture to mimic a situation in which I only saw the perp for a few seconds. I may have a horrible memory, but I didn't do well at all. To have this kind of evidence be taken as absolute truth in a court of law with no other supporting evidence would be tragic and unfair.

http://www.quickflashgames.com/games/eye-witness/

http://www.stumbleupon.com/su/1h3Ypm/www.brainmysteries.com/research/Recognition_of_facial_expressions_is_not_universal.asp

While we were talking about memory and recalling memories from a crime I began thinking about how people tend to remember facial features differently.  This lead me to the question of how do different races affect the recognition of facial expressions?  Do people of different races really notice different facial features on a perpetrator that may lead them to the conviction of an innocent person? 

 I stumbled upon this article that discusses the recognition of facial expressions between Caucasians and Asians. Research showed that both races looked at different facial features when examining facial expressions.  Caucasians tend to look at the shape of the eyes and mouth, while Asians tend to focus only on the shape of the nose. This article gives an insight on what features people focus on when looking at a person.  I believe this will definitely impact their judgment on who committed a crime when looking at a lineup of suspects. 

In class, we have been talking about memory and how witness identification is a key role in many cases. I came across this website, which has a couple face recognition tests. It was really interesting to see that my memory isn't as good as I thought it was. Memory, as defined in class is the ability to recover information and keep it stored safely. These tests mostly focus on short term memory. Try one of the three tests, its shocking to see how good, or bad your memory is!

http://www.faceblind.org/facetests/index.php

 

"Caucasians and Asians don't examine faces in the same way, according to new research. PhD student Caroline Blais, of the Université de Montréal Department of Psychology, has published two studies on the subject: one in Current Biology and the other in PLoS One."

http://www.stumbleupon.com/su/1h3Ypm/www.brainmysteries.com/research/Recognition_of_facial_expressions_is_not_universal.asp 

  Above is a great, albeit quick, example of research being done on facial recognition, and featural processing strategies.  The article suggests that the common assumption that all facial recognition follows a bottom-up approach, in that various individual parts of a persons face are combined and an overall impression is then formed, may be accurate for causasions, but not other races. 

  Using eye-tracking technology the research demonstrated that causasions spend most time shifting between both eyes and the mouth forming a recognizable "triangle" pattern.  However, Asians were demonstrated to spend time focusing on the nose and using a wholistic or top-down approach.  The research went on to demonstrate that Asians also had a difficulty distinguishing between various negative emotions which included subtle facial distinctions (e.g. variations in mouth positioning). 

   I'm interested to know some of the implications of this research to witness identification.  The article suggests that neither bottom-up nor top-down processing is better per se, only that they function differently in various contexts.  It would seem that bottom-up processing should take longer yet be more detailed, however I guess research does not suggest this to be the case.  And what reasons are there for this distinction between Asian and Caucasians facial processing techniques? 

  

   

In my Cognitive Psychology textbook, Cognition: Theory and Applications by Reed, there's a box that talks about why we have trouble distinguishing faces of people who are a different race from us.

People are notoriously awful at recognizing faces from other races. It's a human foible often explained by the notion that we have more experience looking at members of our own race and thus acquire "perceptual expertise" for characteristics of our own kind.
One influential version of that hypothesis argues that the so-called cross-race recognition deficit can be modeled by assuming that faces of other races are more psychologically similar than are faces of one's own race. But Daniel Level, PhD, a cognitive psychologist at Kent State University, has been unsatisfied with that argument.
"The perceptual expertise position is pretty intuitive, and it makes sense," he says. "But I'm arguing that it's not really the case. The problem is not that we can't code the details of cross-race faces - it's that we don't."
Instead, he says, people place inordinate emphasis on race categories - whether someone is white, black, or Asian - ignoring information that would help them recognize people as individuals. In recent research, Levin has shown that people can, in fact, perceive fine differences among faces of people from other races - as long as they're using those differences to make race classifications.
For example, Levin explains, "When a white person looks at another white person's nose, they're likely to think to themselves, 'That's John's nose.' When they look at a black person's nose, they're likely to think, "That's a black nose."
The results are important, Levin maintains, because they help explain the long-standing question of why people are poor at recognizing the faces of people who  belong to other racial groups. Such an understanding could be useful in a variety of settings, including training police and others in the justice system to identify faces more accurately.

Source: From "Why do 'they all look alike'?" by Siri Carpenter, Monitor on Psychology, December, 2000, p. 44.

Facial recognition is of vital importance in many criminal cases.  So much of what lawyers rely on are eye-witness accounts and many of those accounts include facial recognition.  Though there many factors in a person recognizing a face like lighting, partial vision and whether or not  a person was planing one recognizing a face, a new factor has been found in genetics.  This weeks 60-Second Psych podcast introduces and somewhat explains that facial recognition may be genetic.  I thought this was very interesting and thought provoking.  And though they cite that this finding might give light to dyslexia or William's syndrome, I think this may eventually help the psychology & law community.  

This is a facial memory test that shows how well you can remember faces.  This test starts out by showing you a face and you have to guess from three faces which one was the original face.  The test then starts to get harder and harder by showing you six faces for 20 seconds and you have to pick which face was in the lineup of six.  It tn gets harder by putting colors into the faces so you cannot see the details very well.

I got a 76 percent on the test.  It said that a normal person would be able to get around 80 percent.  I think this is a good way of showing how difficult remembering faces can actually be.

http://www.faceblind.org/facetests/fgcfmt/fgcfmt_intro.php 

 

This article reports on research of digitally altering lineup photographs to add or remove distinctive facial features. The reason behind this is that if the witness or victim recalls a distinctive facial feature such as a scar, there is a tendency for them to focus just on that feature. So, if the police pick up an innocent person who has a similar feature, the eyewitness is likely to mistakenly identify the innocent person on the basis of this feature. This technique would either remove the scar from all of the photos in the lineup or add a scar to all of the photos in the lineup.  It turns out that adding the same distinctive feature to each of the faces led to more accurate performance than removing the distinctive feature. People correctly identified the face they had seen about 50% of the time when all of the faces had the distinctive feature, but they identified it correctly only about 30% of the time when the distinctive feature was removed from the target face.

He had a big scar

Here is a little game that I found that demonstrates how difficult it is to recall a face that you have seen for minimal time.  Though the game is a little silly, I think that it does a good job making you aware of how hard it would be for a witness to correctly recall a face in a line up, when trying to explain the criminal in order to construct a composite, or picking out a face in a mug book. 

http://www.stumbleupon.com/su/2MaIuV/faculty.washington.edu/chudler/java/facemem2.html