Resolved NY - Queens, Living Male, Hispanic/Latin under 40, nonverbal, Oct'17 - Name withheld

[Truman]

@houndpounder @Truman

In the past few weeks, along with other research, I've been looking through NamUs Missing Persons, too. In addition to comparing demographic information, circumstances, dates, and photos by visual appearance, I've also utilized a few programs which break down and analyze key points of facial structure / composition, giving a mathematical / algorithmic comparison between two faces, and in some cases (depending on the program and how well technology cooperates at the given time) the likelihood of two images being the same person. I should point out these have not been official programs, and some of the results have seemed 'interesting' to me.

The MP case brought up was one I'd considered. According to the notes I have at hand at the moment, results for comparison of the UIDLI with Edgar Diaz-Cabanas (NamUs #MP39754) yielded possibilities / similarities of 42% and 48%, with possible 'matches' starting at higher percentages (thus, the likelihood of being the same person was not very high).

Truman brings up a great point. Regarding facial or body markings such as moles, there are a few things to keep in mind. (Apologies for not having my sources handy). One thing is that, restricted to nature, these markings are likely to develop but not disappear -- i.e., an earlier picture might lack them, while a later photo might have them. Secondly, such markings may be surgically removed (which yields results basically the opposite of the first note). Third, it's important to know what kind of mark one is looking at. Scars may fade; an acne blemish may heal quickly; a malignant growth might change. One source I read included anecdotes about looking for someone, including and excluding potential candidates based on a mark that wasn't what they thought (for instance, mud or a photographic artifact).

You've made many excellent points in your last few posts...thank you for the time and thought you have put into this. I'll keep working my way through NamUs and Charley with your points in mind, please keep us updated with any new information or ideas. Thanks again!

 

[snakes]

(Here is an example of some of the NamUs file comparisons I've been working on).

Based on the UIDLI's NamUs file information, one of the Missing Persons searches I've entered yielded four results.

The parameters were: current age = 18 - 40; male; Hispanic / Latino; date of last contact = before 17 November 2017; state = New York. (I give some space for d.o.l.c.; I omit height and weight here, in case of significant lapse between dolc and found; hair colour for same reason, because this can easily be changed).

The four results are:

1. Steven Sanchez -- (no image available for comparison); current age 31; missing from Buffalo since 26 december 2016.
The National Missing and Unidentified Persons System (NamUs)

2. Yoxy Zuniga Romero -- missing from Yonkers since 25 December 2016; current age 19.
The National Missing and Unidentified Persons System (NamUs)

3. Erol Melendez -- missing from Dobbs Ferry since 22 December 2016; current age 20.
The National Missing and Unidentified Persons System (NamUs)

4. German Crisoforo Rosales Salinas -- missing from Mamaroneck since 27 November 2008; current age 40.
The National Missing and Unidentified Persons System (NamUs)

According to the three facial comparison programs I have at hand (excluding search result 1): The highest percentage of similarity is person 3 -- according to all three programs and as an overall average; two programs reached percentages near possibility of same person but still fell short. The second highest percentage of similarity was person 2, in all three programs and overall average.

Issues with the comparisons include: date of photos is not always known (differences in age may alter results); differing angles might yield relative inaccuracies; some images are of such a quality as to be found unreadable by one or more programs.

{As a side note: I find it interesting that three of these four persons are missing since a similar date in November (the fourth person missing in december, which is relatively close -- albeit several years before the other individuals), and that three of the four individuals are missing from Westchester County. Two of these persons share both similar dates and the same county.}

[Edit: attempted to reword last paragraph for clarity]

 

[snakes]

{As a side note: I find it interesting that three of these four persons are missing since a similar date in November (the fourth person missing in december, which is relatively close -- albeit several years before the other individuals), and that three of the four individuals are missing from Westchester County. Two of these persons share both similar dates and the same county.}

Oops. Apologies. Edit window has closed. That *should* say: Three of the four are missing since late *december* of the same year, with the fourth missing since late *November* of a previous year. I apologize for switching the months in the original post.

 

[snakes]

Supplement to previous post, regarding the facial comparison programs I've utilized:

With the addition of a fourth program (similarity ranking: 4, 2, 3), the ranking of similarity by averages change to 4, 3, 2 (from 3, 2, 4).

This is to show how the programs can differ from one another, and how one differing program may affect overall average results.

The greatest variance in this example was for the image of person 3 (Erol Melendez), which yielded a difference of around 47 percentage points between two programs, for the same image.

 

[Sleuth91]

Supplement to previous post, regarding the facial comparison programs I've utilized:

With the addition of a fourth program (similarity ranking: 4, 2, 3), the ranking of similarity by averages change to 4, 3, 2 (from 3, 2, 4).

This is to show how the programs can differ from one another, and how one differing program may affect overall average results.

The greatest variance in this example was for the image of person 3 (Erol Melendez), which yielded a difference of around 47 percentage points between two programs, for the same image.

@snakes Thank you for the time that you have dedicated to this case. Awesome, awesome work. May I ask which programs you are using? I'd like to look into them for a case I'm researching.

 

[Giuc0]

I was thinking about the possibility of the UID being First Nations rather than Hispanic because of a case where the UID was thought to be Black, but when identified, the missing person matched by DNA was actually Asian. I don't remember the specific case but I recall the racial disrecpectancy very clearly.
And our UID has traits of First Nations.
I can not exclude that our UID may not be Hispanic, which actually broadens our researching panel.

Here, we have a photo of a living person, which makes me much more cautious about a possible race disrecpectancy/misidentification.
Another case when ME estimated female UID as "white", but the DNA analysis by Parabon resulted that she was Black of Somali descent.
So, we were searching the wrong match based on racial misidentification.

So, I would not write off a potential match on race disrecpectancy alone.
Of course, if you also have "Dentals: wisdom teeth available" on a 40ish UID in Dec 2019 and "wisdom teeth removed antemortem" for a 30ish MP in December 2019, then it's an obvious rule-out!
But if the only difference is race and the other elements fit, then I think that it's wise to consider it.

That's why I think that we shall not restrict our research to Hispanic, but we shall broaden to include also First Nations and Asian.

 

[Truman]

(Here is an example of some of the NamUs file comparisons I've been working on).

Based on the UIDLI's NamUs file information, one of the Missing Persons searches I've entered yielded four results.

The parameters were: current age = 18 - 40; male; Hispanic / Latino; date of last contact = before 17 November 2017; state = New York. (I give some space for d.o.l.c.; I omit height and weight here, in case of significant lapse between dolc and found; hair colour for same reason, because this can easily be changed).

The four results are:

1. Steven Sanchez -- (no image available for comparison); current age 31; missing from Buffalo since 26 december 2016.
The National Missing and Unidentified Persons System (NamUs)

2. Yoxy Zuniga Romero -- missing from Yonkers since 25 December 2016; current age 19.
The National Missing and Unidentified Persons System (NamUs)

3. Erol Melendez -- missing from Dobbs Ferry since 22 December 2016; current age 20.
The National Missing and Unidentified Persons System (NamUs)

4. German Crisoforo Rosales Salinas -- missing from Mamaroneck since 27 November 2008; current age 40.
The National Missing and Unidentified Persons System (NamUs)

According to the three facial comparison programs I have at hand (excluding search result 1): The highest percentage of similarity is person 3 -- according to all three programs and as an overall average; two programs reached percentages near possibility of same person but still fell short. The second highest percentage of similarity was person 2, in all three programs and overall average.

Issues with the comparisons include: date of photos is not always known (differences in age may alter results); differing angles might yield relative inaccuracies; some images are of such a quality as to be found unreadable by one or more programs.

{As a side note: I find it interesting that three of these four persons are missing since a similar date in November (the fourth person missing in december, which is relatively close -- albeit several years before the other individuals), and that three of the four individuals are missing from Westchester County. Two of these persons share both similar dates and the same county.}

[Edit: attempted to reword last paragraph for clarity]

I apologize for being so late with this reply, I replied a few days ago and my links were a mess, so the moderator kindly deleted at my request.

To me, the resemblance between our UID and @snakes ' #4, German Crisoforo Rosales Salinas, is striking. I know hairlines can change over time, but they are so similar in this case. Also, please look closely at each man's eyebrows (on your right as you face the screen in their current NamUs photos, links below.) They both seem to have a VERY similar linear scar in the same eyebrow. Images can be reversed, but the scar seems to be in the same eyebrow relative to the hairline. I apologize because I'd love to post a side-by-side but I am technically inept! German also satisfies my concern about the moles present on the face of our UID. Again, moles are not a rock-solid trait by any means, but his propensity to have moles, in combination with other attributes, checks off another box for me. My only concern is his age, our UID looks younger than German's current age of 40, in my opinion. German is not listed as an exclusion on our UID's NamUs.

German Crisoforo Rosales Salinas, NamUs #MP2405

Our Unidentified Living Male, NamUs #UP57512

 

[CaseyLPC]

The UIM and German definitely resemble greatly, but the UIM has a large mole/discoloration beside his nose near the flaring of the nare and one on his upper cheek below his eye that German does not have. Definitely close, but I don't believe they will end up being a match. that being said, I would still probably call it in as a possibility/ruleout.

 

[Goecke]

Looking at the picture of him, I also agree that it looks like he has vitiligo. Scars can result in pigment loss as well, but there's typically a shine to them that isn't visible in the pics.

How prominent vitiligo is depends 1000% on the person's skin tone. Since vitiligo is pigment loss, it's less or more prominent depending on the amount of pigment someone has in their skin to begin with. The affected areas are usually around the same colour no matter what the person's skintone is- Therefore it's a lot more noticeable on people with darker skin. Since our UIM has relatively light skin, it makes sense that it's not as obvious on him.

 

[Giuc0]

Looking at the picture of him, I also agree that it looks like he has vitiligo. Scars can result in pigment loss as well, but there's typically a shine to them that isn't visible in the pics.

How prominent vitiligo is depends 1000% on the person's skin tone. Since vitiligo is pigment loss, it's less or more prominent depending on the amount of pigment someone has in their skin to begin with. The affected areas are usually around the same colour no matter what the person's skintone is- Therefore it's a lot more noticeable on people with darker skin. Since our UIM has relatively light skin, it makes sense that it's not as obvious on him.

The vitiligo has been confirmed in his records. Source: LE


NY - NY - Queens, Living Male, Hispanic/Latin, under 40, black hair, brown eyes, nonverbal, 9 Oct 2017
My post where I called LE exactly for this

 

[Giuc0]

I apologize for being so late with this reply, I replied a few days ago and my links were a mess, so the moderator kindly deleted at my request.

To me, the resemblance between our UID and @snakes ' #4, German Crisoforo Rosales Salinas, is striking. I know hairlines can change over time, but they are so similar in this case. Also, please look closely at each man's eyebrows (on your right as you face the screen in their current NamUs photos, links below.) They both seem to have a VERY similar linear scar in the same eyebrow. Images can be reversed, but the scar seems to be in the same eyebrow relative to the hairline. I apologize because I'd love to post a side-by-side but I am technically inept! German also satisfies my concern about the moles present on the face of our UID. Again, moles are not a rock-solid trait by any means, but his propensity to have moles, in combination with other attributes, checks off another box for me. My only concern is his age, our UID looks younger than German's current age of 40, in my opinion. German is not listed as an exclusion on our UID's NamUs.

German Crisoforo Rosales Salinas, NamUs #MP2405

Our Unidentified Living Male, NamUs #UP57512

We had estimated ages off by decades with unidentified.
Excuse me for not remembering exactly the case, but the UID was estimated to be in their 30s and the DNA match was a 16-20-something.

So, if it's your only con, I would not worry too much.

 

[snakes]

Greetings, all. Special thank you to those who have continued following along. It's encouraging to see continued interest in helping this individual, and support for each other who are working toward that end (whether by researching, posting, or following along). There are several new posts, and I do have forthcoming replies, if I may. My time to post here is somewhat sporadic and irregular, and I'm not sure if I'll get to all of them today. I appreciate both the attention being kept on this individual's case, and patience with my contributions. Thank you.

 

[snakes]

@snakes Thank you for the time that you have dedicated to this case. Awesome, awesome work. May I ask which programs you are using? I'd like to look into them for a case I'm researching.

@Sleuth91. Thank you for your kind words and continued interest in, and attention to, this UIDLI's case. It's great to log in and see familiar names.

Regarding facial comparison programs:

There's quite a bit to be said on the subject, and I wonder if there's a more proper Forum in which to more thoroughly discuss the topic.

Based on an abundance of caution regarding rules, and also considering the mixed results I've experienced thus far, I wouldn't feel very comfortable recommending any specific programs at this time. (Some sites / apps might feature pop-up ads &/or advertise further services or features for payment). Depending on one's available financial and technological resources, there seems to be a range of programs available to the general public, on websites and as applications, which (at least claim to) analyze and/or compare images of faces. Again, based on my own experience, I'd advise caution. Maybe the best way I could put it would be to think of them as another WebSleuths user's opinion, and not necessarily as the teacher's answer key. That's my current opinion, and I imagine programs tend to vary.

If you can think of any other way I might be able to help on the case you're researching, please feel free to let me know. May you contribute well and may the case be resolved as best it may.

 

[snakes]

I was thinking about the possibility of the UID being First Nations rather than Hispanic because of a case where the UID was thought to be Black, but when identified, the missing person matched by DNA was actually Asian. I don't remember the specific case but I recall the racial disrecpectancy very clearly.
And our UID has traits of First Nations.
I can not exclude that our UID may not be Hispanic, which actually broadens our researching panel.

Here, we have a photo of a living person, which makes me much more cautious about a possible race disrecpectancy/misidentification.
Another case when ME estimated female UID as "white", but the DNA analysis by Parabon resulted that she was Black of Somali descent.
So, we were searching the wrong match based on racial misidentification.

So, I would not write off a potential match on race disrecpectancy alone.
Of course, if you also have "Dentals: wisdom teeth available" on a 40ish UID in Dec 2019 and "wisdom teeth removed antemortem" for a 30ish MP in December 2019, then it's an obvious rule-out!
But if the only difference is race and the other elements fit, then I think that it's wise to consider it.

That's why I think that we shall not restrict our research to Hispanic, but we shall broaden to include also First Nations and Asian.

@Giuc0 There doesn't seem to be any post that has said only persons stated to be of Hispanic/Latinx racial/ethnic background should be considered, but it might be a good idea to overtly clarify this, as you have.

For me, I've definitely included in my research all sorts of possibilities, including typographical errors or well-intentioned, inadvertent misinformation -- I'm sure sometimes the information or presumptions that Law Enforcement is working with, for example, are only as good as they can be. I imagine we've all made mistakes in life, despite our best efforts. Sometimes the given information is more or less reliable, and sometimes what we have is all that could be released. These are my opinions.

Personally, I would not rule out any individual based solely on ethnic/racial disparity, at least for this case.

 

[Dylansmom08]

Sad to see that he is still unidentified I have none many searches since first seeing this post and have come up with no promising leads so far. I will keep looking for clues and answers though. I pray that while he is waiting to be identified he is being well taken care of in these difficult times of the pandemic.

 

[snakes]

I apologize for being so late with this reply, I replied a few days ago and my links were a mess, so the moderator kindly deleted at my request.

To me, the resemblance between our UID and @snakes ' #4, German Crisoforo Rosales Salinas, is striking. I know hairlines can change over time, but they are so similar in this case. Also, please look closely at each man's eyebrows (on your right as you face the screen in their current NamUs photos, links below.) They both seem to have a VERY similar linear scar in the same eyebrow. Images can be reversed, but the scar seems to be in the same eyebrow relative to the hairline. I apologize because I'd love to post a side-by-side but I am technically inept! German also satisfies my concern about the moles present on the face of our UID. Again, moles are not a rock-solid trait by any means, but his propensity to have moles, in combination with other attributes, checks off another box for me. My only concern is his age, our UID looks younger than German's current age of 40, in my opinion. German is not listed as an exclusion on our UID's NamUs.

German Crisoforo Rosales Salinas, NamUs #MP2405

Our Unidentified Living Male, NamUs #UP57512

Hello, @Truman It seems your sense of punctuality is more refined than mine; I hadn't been back until yesterday. No worries about wherever you are in your technological proficiency -- we're all somewhere. There are many skills to have in life, and that is just one. Sometimes it takes time to get a knack; sometimes it's a matter of matching the right tool to the right person. I'm glad the moderator was helpful, and that you've posted again.

Regarding German Crisoforo Rosales Salinas: I agree there may be some resemblance. There is definitely a lot more for me to learn about analyzing and comparing faces; I can't really add much to your assessment at this time. I've found some potential information about hair lines (which was a subject I had been recently pondering); I hope to look through it soon. There does seem to be a similar scar with those two. I think I have seen a similar marking on at least one other missing person, but don't recall who, at this time. Side-by-side comparisons are something else at which I've been looking. One issue with them has been that the image of 'our' UIDLI has their head slightly tilted. The program I've been using only rotates by a particular number of degrees, larger than what is here needed. I was able to capture a 'corrected' image, from a different program, but the size is small and quality is lost when enlarging. If you would like, I can post, if I may, either 'tilted, with somewhat better quality' or 'upright, with somewhat blurred quality'. (If such images would not be against the WS rules). Personally, I've been preferring to keep the UIDLI on the left, and any potential match on the right, lately, but I could re-orient those for you if you'd prefer the other way around. The presence of nevi ('moles', 'beauty spots', etc.) does seem to make for a closer match than some others; I agree, though, that that is not the only factor which should be considered. As for age, I have a few thoughts -- some of us look considerably older or younger than our legal age; we're still not sure when the UIDLI's picture was taken (it could be from 2017 -- there hasn't yet been any confirmation on actual nor relative date -- which I concede is not very long ago, but might make some difference); it is my understanding that sometimes ages might not be exact in missing persons reports, depending on who reports, what information may or may not be found in official systems, and which information law enforcement chooses to divulge (allowing for confirmation by withholding some info). I might revisit the NamUs files again. The last time I had checked, there had been no exclusions listed for the UIDLI. Thank you for your contributions. [Edit: typo].

 

[Giuc0]

@Giuc0 There doesn't seem to be any post that has said only persons stated to be of Hispanic/Latinx racial/ethnic background should be considered, but it might be a good idea to overtly clarify this, as you have.

For me, I've definitely included in my research all sorts of possibilities, including typographical errors or well-intentioned, inadvertent misinformation -- I'm sure sometimes the information or presumptions that Law Enforcement is working with, for example, are only as good as they can be. I imagine we've all made mistakes in life, despite our best efforts. Sometimes the given information is more or less reliable, and sometimes what we have is all that could be released. These are my opinions.

Personally, I would not rule out any individual based solely on ethnic/racial disparity, at least for this case.

I was answering to this post, but I forgot to quote when I typed my answer: NY - NY - Queens, Living Male, Hispanic/Latin, under 40, black hair, brown eyes, nonverbal, 9 Oct 2017

The poster disliked racialising based on traits. I wanted to clarify that I didn't want to racialise because I thought "Hispanic folks must be like this or that", but I raised the issue because racial misidentification by LE/ME brought us on wrong possible matches.

I know that the cases I remember were UID and he is a living gentleman, but I wanted to raise the issue of potential racial misidentification by LE. However, I should had stated it loud and clear right off the bat in my first answer for avoiding misunderstandings like these. My bad!
I should learn from this mistake.


ETA: First Nations can be misidentified as Caucasian, Asian, Hispanic. And vice-versa.
Asian, especially South Eastern Asian, can be misidentified as Black. I don't remember exactly the case and I am too lazy to search right now, but I remember clearly that it indeed happened.
Websleuths taught me about this issue when I dived into the subject. I wanted to remind what may be obvious but can be easily forgotten/overlooked with all the details to cross-check for finding a match.
To make an example, when my cousin travels to Mexico, she is often identified by Mexican folks in the streets as of First Nations. She was born and raised by a Sepharadi Jewish father and a Bearnese mother (Bearn is a region in the South Western France, near the Spanish border) and lived near Pau since she was born until adulthood. She's been to Mexico for the first time only at 25yo. My cousin being misidentified as First Nations was told by my mom and my cousin herself. She is alive and well. I give this example to highlight how easy is racial misidentification!
Such misidentifications are so common that for a living unidentified, I want to raise this bright red flag right off the bat, so we avoid tunnel vision and wrong tips.

 

[Bit of hope]

I wonder how he's doing now?

Love this question! First thought I had coming here.

 

[Bit of hope]

Just a few remarks. This poor man was in a hospital in October 2017. We don't know if he recovered or has died in the mean time. It's 2020. Where are the updates? (sorry, I didn't read the whole thread) Is he still unidentified? To answer my own question: No he isn't identified because he wouldn't have been in NamUs anymore....but sorry....why isn't there an updated photo? That is ridiculous. And BTW any person who hasn't the moles can be ruled out IMO or we should go and look for persons who were very young when missing..and the moles still weren't there.

edited by me a lot of times...sorry....

 

[Bit of hope]

I searched for him on Guatemala missing person sites, on FB (Garcia related families), even went back as far as Peru (oh dear....it made me so sad....the year is 2020 and there is no real missing persons system, it heartbreaking, like a lot of other South American countries) It's very hard for me to go on looking out for him....so many frustrations....and there is nothing new on his case....BUT, every person has the right to get his/her name back.