Sunday, October 11, 2015

Forensic Examiners Pass the Face Matching Test

Forensic Examiners Pass the Face Matching Test

Yumi Mita

"Forensic Examiners Pass the Face Matching Test." ScienceDaily. ScienceDaily, 2 Sept. 2015. Web. 11 Oct. 2015.

Are the law enforcers, who have been trained for facial recognition, good at their job as we expect them to be? The answer is, yes. A study to test the skills of facial forensic examiners have been performed, and the research suggests that the examiners identify better than the average person or computers, while using analytical methods. The facial experts’ use of analytical methods leads to an understanding that they identify faces differently from the super-recognizers, a group of people who are naturally good at face-matching.
The study, which is published in the Proceedings of the Royal Society B with help of many scintillating minds, such as colleagues at the National Institute of Standards and Technology and the University of Texas at Dallas in the US, was tested on 27 international, experienced facial forensic examiners who were attending a meeting of the Facial identification Scientific Work Group. These members belong to agencies like the FBI, homeland security, and police and customs of various countries, including the US.

How were they tested? All examiners were given three tests where they had to decide if the same person was pictured in the paired photographs. This may sound like an easy task but the chosen images were particularly challenging, given the fact that “computer algorithms were 100 percent wrong on one of the tests.” The forensic examiners’ performance was then compared to non-experts’ performance, who were the control group of this experiment. The results indicated positive, high performance levels of the experts, but Dr. White, a research leader of UNSW Australia psychology, reminds us that “although the tests were challenging, the images were relatively good quality. Faces were captured on high-resolution cameras, in favorable lighting conditions and subjects were looking straight at the camera,” meaning that harsher tasks are bestowed upon the examiners “when images are extracted from surveillance footage.”

Overall, the article gave a thorough explanation and procedure of the study that highlighted the expertise of facial forensic examiners, which was delightful to hear as a member of a society where crimes are occurred daily. It also included a sample image of what might the test have used, allowing the readers to connect to the study more personally. This article focalized on sharing the result of a study so I think it was just right, without any weaknesses. However, it could have been made more intriguing if an opinion of a test-taker, examiner or control, was integrated in the piece.

16 comments:

Anonymous said...

Yumi wrote a fair review of the article “Forensic Examiners Pass the Face Matching Test”, which describes the study that tested the skills of facial forensic examiners. By opening the review with a question that may have been on the mind of the viewer when they read the title, it captured the attention of the reader. Since one has to read on to determine the answer to this question, this method gives a concise flow to the opening paragraph. Yumi also used this question-and-answer method in the third paragraph, which explained the actual experiment performed. I enjoyed the integration of select quotes into Yumi’s review as well. The usage of the quote, “computer algorithms were 100 percent wrong on one of the tests”, gave exceptional emphasis on this fact that forensic examiners are able to use certain methods for identification that exceed the ability of computers. The additional quote that she uses, which gives an explanation of the type of photographs that were utilized in the experiment, allowed the reader to understand an important flaw in the study. This flaw was that the photos were of fair quality, with the the subject looking directly at the camera, and good lighting conditions, which, in reality, may not be available to the examiner. Lastly, the flow of all of Yumi’s paragraphs was flawless. The opening of the review was a general statement of the results of the study, which led to the details of the experiment (such as who was involved), and ultimately led to the details about the actual study and how the forensic examiners were tested.
In order to make this review even more solid than it already is, as a reader, I would have welcomed an explanation about the types of analytical methods that the examiners used in order to determine if the two pictures were of the same person. Also, I was unsure of who the “super-recognizers” were in relation to the article and review. Was this group of people who are “naturally good at face-matching” the control group of non experts in the experiment? This fact was not as clear to the reader. Otherwise, Yumi was able to write an excellent review for this article and was able to capture the essence of what this study is worth to the forensic community.
I was especially impressed by the fact that the forensic examiners tested were able to complete a task that “computer algorithms were 100 percent wrong on…” Also, I was amazed by the skill of these examiners who could “identify better than the average person or computers”.

Anonymous said...

The blog post did a good job at presenting the main points of the article without complicating the subject and dragging out storyline. It was easy to makes sense of how the article was trying to portray the talent and skill of those trained for facial recognition. The blog post also did a good job at showing the reader how forensic scientists are tested and proven to be more efficient than mechanical facial recognizers. This was shown when testing three forensic scientists on identifying whether two pictures of a face were the same person or not, and then comparing their conclusions to a completely unskilled group of people used as the “control group.”
I found that the review could have been a little more interesting if it had talked about what forensic scientists do much better when identifying a face compared to a machine. It would be interesting to know how a machine works in this situation compared to a scientist on their own. Also, I would have like to know more about the testing of the three scientists at Facial Identification Science Work Group and how they worked to decipher whether or not the face they were examining was, in fact, the same person or not. I am also skeptical of whether those tests are valid since the pictures they were given were in such good quality and staged, rather than being from a real life situation.
I was very impressed when reading that the work of forensic scientists is more reliable for facial recognition than the work of a machine. Almost everything we do today can be done in a quicker and better way when a machine is doing the work, so it was surprising to hear that there are still some things that an individual can do better on their own than with the help of technology.

Anonymous said...

The blog post did a good job at presenting the main points of the article without complicating the subject and dragging out storyline. It was easy to makes sense of how the article was trying to portray the talent and skill of those trained for facial recognition. The blog post also did a good job at showing the reader how forensic scientists are tested and proven to be more efficient than mechanical facial recognizers. This was shown when testing three forensic scientists on identifying whether two pictures of a face were the same person or not, and then comparing their conclusions to a completely unskilled group of people used as the “control group.”
I found that the review could have been a little more interesting if it had talked about what forensic scientists do much better when identifying a face compared to a machine. It would be interesting to know how a machine works in this situation compared to a scientist on their own. Also, I would have like to know more about the testing of the three scientists at Facial Identification Science Work Group and how they worked to decipher whether or not the face they were examining was, in fact, the same person or not. I am also skeptical of whether those tests are valid since the pictures they were given were in such good quality and staged, rather than being from a real life situation.
I was very impressed when reading that the work of forensic scientists is more reliable for facial recognition than the work of a machine. Almost everything we do today can be done in a quicker and better way when a machine is doing the work, so it was surprising to hear that there are still some things that an individual can do better on their own than with the help of technology.

Anonymous said...

In Yumi Mita’s review of the article “Forensic Examiners Pass the Face Matching Test”, she talked about if law enforcers have been trained for facial recognition and how good at their job do they have to be in order to do this successfully. She said that the law enforcers have been tested for their skills as forensic examiners and they have performed better than the average person or computers. She also states in her review that the article included what the facial experts use. They use analytical methods that leads to a better understanding on how to identify faces differently from the super-recognizers. She talks about how a study was published and that it included many scintillating minds and she went into description about how there was a meeting of the Facial Identification Scientific Work Group that had members including agencies like the FBI, homeland security, and police and customs of various countries, including the United States of America. An area where Yumi could have focused more on, is what it takes to be a good face identifier as a law enforcement officer. She also could have included more detail on what the test had included in order for the law enforcement officers to pass the test and what would have happened if they did not pass the test. I have learned a lot from reading her review. I always thought that face-identifiers was from the computer, but I learned that it can be from people as well. I still don’t completely understand how a person can identify someone by looking at them, but I am going to be looking more into the topic to understand it completely. It has changed my impression of what it takes to be in law enforcement because you have to go through all these tests to prove that you are good at your job or you have a good understanding of the work that you need to do.

Anonymous said...

Yumi, I really enjoyed reading your review of the article “Forensic Examiners Pass the Face Matching Test. There were many aspects of your review that I thought were great. First, I liked how you started your review with a question. This question made me want to read on further. Second, I liked how your review was straight to the point. You added the right amount of information that I understood the topic without overwhelming me. Third, I liked how you added a quote from a scientist in this field. This quote increases the strength of your review. Although I liked your review, there are a few things you could do to make it even better. One, you talk about how examiners are better than machines, but if you could go into more depth and say what exactly they are better in would be great. Second, I was a little confused about the super recognizers. Maybe if you went into a little more depth I would have understood them more. However, maybe the article didn’t explain it, which is why you didn’t explain it. One thing I was really impressed with was that forensic examiners could use analytical methods better than a machine. You would think that a machine would clearly be better. Again, I really enjoyed reading your review.

Anonymous said...

The blog post I read was Yumi Mita’s review of the article, “Forensic Examiners Pass the Face Matching Test.” Yumi discussed how several facial forensic examiners from around the world were tested on how accurately they could do facial recognition and whether the examiners skills were greater than that of a computer system or even just a regular group of people. Some of the examiners came from agencies including the FBI, homeland security, and several different police stations. I feel that Yumi did a great job of explaining what the study consisted of and how it was carried out. She went into detail on what the facial examiners had to complete as well as gave an in depth review of the studies results. I do think that the blog post could have gained more appeal if more information was added about what facial examiners do and what sort of cases they solve. I also think she could have better explained why the study was done and who came up with the idea to do this study. Yet, she did bring up the argument that was made which stated that the photos used in the study had perfect quality and good lighting, while for many of the cases in which the facial examiners are needed the photos are from surveillance cameras and of people far away. Overall, I feel that Yumi did an excellent job expressing what the study consisted of and how facial examiners are in fact good at their jobs.

Anonymous said...

Yumi did a solid job of reviewing this current event article. I especially liked how the opening line of her review is a question to her audience, rather than just a statement that provides background information of her article. By asking this question, and then providing the answer based on her article, she captures the audience’s attention because it feels more engaging. I also think that Yumi did a good job of explaining why facial experts are more reliable than computer analyzers-- which, in all, was the point of the article. It was nice that Yumi gave a brief introduction about testing the experts before she went into describing how the experts were tested. She did a very good job of describing to her audience how the experts were tested because it was easy to understand. I also appreciate her use of direct quotes, like how she quoted Dr. White on his evaluation of the experiment. This not only provided more insight into what the experiment was like, but also it established some credibility in Yumi’s review.
Overall, Yumi’s review was very good but she did make a few grammatical errors that slowed down the fluidity of the review. They were very minor, however, without them, the review would be completely comprehensible and the reader would not have to reread certain lines twice. I also wished she had explained the difference between facial experts and super-recognizers a little bit more in depth, because I found that sentence to be a little unclear.
In general, I was surprised that a study such as this one had been done so recently. I assumed that an experiment that determines the accuracy of facial forensic examiners would have been performed a lot sooner, considering that these scientists are heavily relied on in criminal investigations. Lastly, I also agree with Yumi that the article would have been more interesting if the author incorporated the opinions of test-takers and researchers in order to provide more insight.

Anonymous said...

Emme Kerj
I really liked that you right away acknowledged the fact that some people may not think that facial recognition law enforcers have legitimate skills since it is easy to assume that computers would be able to do a better job. You also explained very thoroughly how the scientists were tested so there was no confusion about how it was concluded that facial recognition scientists are greater at their job than anyone/anything else.
One thing that the article mentioned was that the increased use of CCTV and images captured on mobile phones has made facial recognition scientists to become extremely important in identifying subjects and I would have liked to read about that in your review since it kind of explains why the article was written in the first place. I also think that you could have furthered your opinion on why it is so important to have facial recognition scientists and how they can help during investigations. I also think that you could have put a little bit more emphasis on the possible sources of error that the test the scientists took had and why it is hard to always be accurate when identifying suspects in photographs and videos.
Something that surprised me was that the results showed that these expert facial recognition scientists were better at pairing individuals than computers because I would have absolutely assumed that a programmed computer would be able to do a better job.

Anonymous said...

Ian Baxter
I thought that Yumi did a great job reviewing the article posted on sciencedaily.com. One of the many things I believe she did particularly well with in regards to her review was integrating quotes directly from the article. This made the review more concise and easy to understand, as the reader of her review did not have to go through the article numerous times searching relentlessly for answers. Another thing I thought Yumi did a great job with was explaining the expertise required of facial forensic examiners. She constantly mentioned that even computers would get matches wrong that the examiners would accurately guess. “Computer algorithms were 100 percent wrong on one of the tests”. Finally, I thought Yumi did a fantastic job when it came to grasping the audience’s attention. The hook at the beginning of her review came in the form of a question, and instantly I was intrigued on reading more. Throughout the review she did a great job maintaining the audience’s interest in the review.
While Yumi did numerous things very well on her review, one of the things she could have improved on in it was discussing the subject’s relevancy to society a little bit more. She did a great job at analyzing and describing the important points of the article but I felt that if she had discussed the importance of facial recognition a little bit more, the review would have been more whole. Another thing that Yumi could have improved on was discussing the difference between facial forensic examiners and super-recognizers. This part of the review was very intriguing and I believe that if Yumi went into more depth regarding this subject I would have better understood what a super-examiner actually was and what role they play in forensic science.
One thing I was impressed by after reading the article and its review was that facial forensic scientists are deemed more accurate than machines. Its typically the other way round such that machines are more accurate than humans, so it was interesting to see that humans are better at facial recognition than machines.

Anonymous said...

I liked how you introduced a question which the article presented and then proceeded to explain how the article answered it. You gave great detail of the tests preformed but some of the information was hard to understand. You gave example of some terms such as "super examiners" and identified them, but didn't fully define them. That would have made it more tangible for us to understand your analysis. I personally thought this article was a very interesting topic, and it was nice to know that experts do a better job of being able to recognize faces than a computer. These days I mostly only hear about how technology can do things more accurately than we can. I think it was great that you included the flaws that the test could have; that the experts usually have to identify more blurred faces. Giving both sides of the argument made your summary more credible. Something that was missing in the article was why this is important into society. Talking about how this could be helpful/ non helpful with the ability to lie on computers in the future would have been a good point to jump off of. Overall I think your review answered a lot of important questions the article was regarding in a very intriguing manner. Great Job.

Anonymous said...

I thought you did a nice job explaining how the people trained for facial recognition are actually very good at their job, when people just expect them to be good without really knowing. You did a nice job getting right to the point. I thought one way you really dragged the readers in was by asking us a question at the beginning of the paragraph. Asking a rhetorical question, then stating the answer gets us, as readers, thinking as we read and keeping the questions in the back of our mind. One way you could have improved your reaction would be by explaining more about the facial experts' and when they are used, or in what cases are they used. Also, explaining any error that the technology could’ve had while running these experiments making it hard to be accurate with the photographs. The topic of the article was very interesting and eye opening. Overall, expanding your ideas would benefit and make your analysis better, but it was a great, informational, relevant, and interesting piece!

Anonymous said...

I thoroughly enjoyed reading Yumi’s review on a new topic that is exciting and fresh information, that helps us as a class learn more about the forensic world. Yumi did a great job o starting off her review with a hook to catch the reader’s eye and setting the scene for an interesting discussion. Yumi also did a great job of thoroughly describing how the process of a facial examiners job works, and how it differs from modern technology. She did a great job of comparing it to modern technology and stating the reasons why good old-fashioned eyes are better than new computers. I think that it could have been beneficial to this review to include more information on the ne technology as well as a tad bit more information of the process of a facial examiners job. I also think that it could have been beneficial to include more information about the job of a facial examiner and how one becomes a facial examiner, and what skills the job requires. All in all this was one of the most interesting current events that I have read, as it was informative and interesting, and taught me about an area that I was not familiar with at all.

Anonymous said...



Yumi did a great job reviewing the article one thing I really enjoyed was her integration of the quotes. The quotes allowed for an easier understanding of the article. She also did an amazing job explaining how important facial examiners are. Especially since, computers get some matches wrong that the examiners would have gotten right. This means that the examiners are absolutely crucial to a case since their testimony is in higher regard than computers. I also really enjoyed how Yumi was able to keep my attention from start to finish.
Although Yumi did an excellent job I wish she had related the importance of facial recognition into society. As well as how important it is that facial examiners are to the criminal process. I would have also like for her to go deeper into super-recognizer. This seems like a super interesting subject and would have completed this excellent review.
The truly interesting thing about this article was that the computers are actually extremely inaccurate and that humans are more likely to correctly recognize the faces.

Anonymous said...

I think that Yumi did a good job at summarizing the article. She did a great job in explaining that law enforcers who were trained for facial recognitions were good at their job. She also made the review straight to the point which helped me understand the topic. Lastly, she included a quote from the article which enhanced the information written in the review. One thing she could have improved was to go into a bit more depth on why it is believed that examiners identify better than computers. Also maybe what kind of mistakes a computer could make that an examiner wouldn’t. Another thing that could have been improved is expand more on who the super-recognizers were. Before reading this review, I thought that since technology is advancing that it couldn’t be replaced, but after reading this review I found out that sometimes technology is replaced by people. So it was a bit shocking that people can do a better job in some things by not using technology. Overall, I enjoyed reading the review and thought it was done well.

Anonymous said...

Yumi did a good job at explaining the why a facial forensic examiner is better at identifying faces then the average person or computers. Yumi explained that facial forensic examiner use analytical methods to identify faces. Yumi also explained the face matching test that the Forensic Examiners take very well. She explained that chosen images were very challenging because even the computer couldn’t recognize them. Also Yumi presented very well that facial examiners have a harsher task because they have to analyze images taken from surveillance footage.
One suggestion I have to improve this review is to go more in detail about what else facial examiners. For instance she said that forensic examiner use analytical methods to identify faces but didn’t explain what the methods they use are. Another suggestion I have is to put more statistics of the result of the test in her review because I only saw one which was “computer algorithms were 100 percent wrong on one of the tests.”
After reading this review I learned that facial forensic examiner are much more effective at analyzing facial recognition then a computer or the average person. This surprised me because in many crime movies and TV shows facial recognition is almost always done by a computer. This review changed my perception on facial forensic examiner because I used to think computers only did facial forensic examinations.

Anonymous said...


Yumi did a great review of the article “Forensic Examiners Pass the Face Matching Test” She explained first off explains how cops are trained for facial recognition. She then goes into depth with a study done at the University of Texas and the National Institute of Standards and Technology. This study was a test done on 27 experienced facial examiners. She then does a great layout of what the test consisted of including how each person had 3 tests , an example of one is where they had to decided if the same person was pictured in the pair photos.

I wish she had compared a person doing facial recognition versus a machine doing so. This could show our growth in technology. This includes what impact does the CCTV has on this experiment. Another thing that I hope Yumi will do in the future is maybe compare the article to something currently happening in society. Would this technique helped some case in the world?

Overall, Yumi did a great job summarizing the article. Facial recognition is the new technology and it is a fascinating topic.