POST 157: USING AI TO CONFIRM THE MISIDENTIFICATION OF REINHARD HEYDRICH, “THE BUTCHER OF PRAGUE” IN POSTS 133, PARTS I & II

 

Note: In this post, I discuss the process I went through to confirm that I’d incorrectly identified the sinister Nazi Reinhard Heydrich in a group photo taken in 1936/1937 at Castle Kamenz (Kamieniec Ząbkowicki, Poland). To prove this to my satisfaction, I made use of an AI-powered tool.

Related Posts:

POST 133: “THE BUTCHER OF PRAGUE,” THE STORY BEHIND A UNIQUE PHOTO OF REINHARD HEYDRICH (PART I)

POST 133: “THE BUTCHER OF PRAGUE,” THE STORY BEHIND A UNIQUE PHOTO OF REINHARD HEYDRICH (PART II)

 

An astute reader recently informed me the person I had identified in Posts 133, Parts I & II, as Reinhard Heydrich (Figure 1), the Nazi’s notorious “Butcher of Prague,” in a group photo taken at Castle Kamenz in Silesia (today: Kamieniec Ząbkowicki, Poland) in 1936/37 (Figure 2) is not him. The picture in question was originally sent to me by a very reliable informant claiming a noted scholar had recognized Reinhard.

 

Figure 1. The individual in the group photo taken at Castle Kamenz (today: Kamieniec Ząbkowicki, Poland) in 1936/37 who I mistakenly identified as the notorious Nazi henchman Reinhard Heydrich

 

Figure 2. The group photo taken at Castle Kamenz (today: Kamieniec Ząbkowicki, Poland) in 1936/37

 

The German gentleman who questioned the identification is a physics teacher in Dresden with an avid interest in history, particularly German military history up to 1918. The man grew up in Gotha in the former German Democratic Republic or East Germany. According to the reader, Reinhard Heydrich’s doppelgänger is Wilhelm Prinz von Hessen-Philippsthal-Barchfeld (1905-1942), and he sent me a link to a photo of him in his military ensemble as well as his bio. (Figure 3) He claims that in the group picture Wilhelm von Hessen, who was head of a “Motor-Sturmbann” of the SS-Standarte 27, is seen wearing his military uniform with the collar tabs “M” for Motor-Sturmbann; even on the highest resolution picture of this gathering, I have difficulty distinguishing the collar tabs.

 

Figure 3. The photo the amateur historian from Dresden sent of Wilhelm Prinz von Hessen-Philippsthal- Barchfeld (1905-1942) who is Reinhard Heydrich’s doppelgänger

 

Since this is my blog the responsibility for fact checking the accuracy of the information I publish in my posts ultimately rests on my shoulders. Given the lengths I’ve gone to caution readers about cloning vital data from other people’s ancestral trees without sourcing the primary documents and verifying their accuracy, I decided I needed to take a similarly rigorous approach in determining whether the amateur historian is accurate.

Let me explain to readers who Wilhelm von Hessen was and whether it is even conceivable he could have been photographed at Castle Kamenz at the time the picture was taken. In 1936/37 Castle Kamenz was owned by Friedrich Heinrich von Preußen (1874-1940) (Figure 4) shown elegantly attired and seated in the front row of the group picture, second from the left. His younger brother was Friedrich Wilhelm von Preussen (1880-1925) (Figure 5) who obviously was not in attendance since he had died prematurely about a dozen years earlier.

 

Figure 4. Friedrich Heinrich von Preußen (1874-1940), owner of Castle Kamenz at the time the 1936/37 photo was taken

 

Figure 5. Friedrich Wilhelm von Preußen (1880-1925), Friedrich Heinrich’s older brother who predeceased him

 

Friedrich Wilhelm was a member of the House of Hohenzollern and a great-grandson of King Frederick William III of Prussia. He was married to Princess Agatha of Hohenlohe-Schillingsfürst from Ratibor & Corvey (1888-1960) (Figure 6), seated as a widow in the center of the front row (Figure 7); together they had four daughters. (Figure 8) The third daughter, Princess Marianne von Preußen (1913-1983) was married to Wilhelm von Hessen. Thus, based on family connections, it is entirely plausible that he attended the family gathering at Castle Kamenz in 1936/37.

 

Figure 6. Friedrich Wilhelm von Preußen and his wife Agatha of Hohenlohe-Schillingsfürst from Ratibor & Corvey (1888-1960) around the time they got married surrounded by family

 

Figure 7. The widow Agatha of Hohenlohe-Schillingsfürst at the time the 1936/37 photo was taken

 

Figure 8. Friedrich Heinrich von Preußen, Princess Agatha of Hohenlohe-Schillingsfürst, and their four daughters as children

 

Obviously, the individuals in attendance at the gathering at Castle Kamenz are unidentified though some are familiar to the reader who furnished the group picture. I’m not entirely positive but think Princess Marianne von Preußen is standing in the second row, the third person from the left dressed in white. Though Friedrich Wilhelm and Princess Agatha’s four daughters were born between 1911 and 1919, as young women they are often difficult to tell apart.

The reader who shared the original photo included side-by-side photos of Reinhard Heydrich and the man purported to be him in the group picture. (Figure 9) Given that Heydrich’s identity had putatively been confirmed, I did not compare the images as carefully as I should have when writing Posts 133, Parts I & II. Had I done so initially, I might have observed a few things that later triggered some doubts. First, I noticed in the picture that is assuredly of Heydrich that he had a narrower face, longer nose, and bigger ears than the individual on the right. Second, and this is much more impressionistic, the person in the group picture is broadly smiling, an attribute I would hardly associate with as sadistic an individual as Heydrich.

 

Figure 9. Reinhard Heydrich (left) and his doppelgänger

 

Trained as an archaeologist steeped in the scientific method, I was still not convinced the person in the group picture was Wilhelm von Hessen. Because the link sent to me by the amateur historian depicting Wilhelm in his military uniform was not conclusive (see Figure 3), I went in search of other pictures of him. Because Wilhelm came from a royal family and married into another royal family, I had the good fortune to find several of them on the Internet and in my own collection of photos.

The first high quality picture I found of Wilhelm von Hessen was taken at the marriage of one of his wife’s sisters, Princess Luise von Preußen (1912-1973) to Moritz Richard Bruno Wilhelm Schmaltz (1901-1983) on the 30th of November 1936, possibly at Castle Kamenz. (Figure 10) It shows Princess Agatha, her four daughters, three sons-in-law, and one grandson. On the very far right in this picture can be seen Princess Marianne von Preußen (1913-1983) and her husband Wilhelm von Hessen. Regrettably, comparing this image of Wilhelm von Hessen to the group photo again yields no definitive answer to the naked eye as to whether it was Wilhelm.

 

Figure 10. From left to right: unidentified person (possibly Princess Marie’s husband, Aloys Rudolf Hug (1885-1972)), Princess Elisabeth von Preußen (1919-1961), Princess Agathe, unidentified grandson, Princess Luise von Preußen (1912-1973), her husband Wilhelm Schmaltz (1901-1983), Princess Marie von Preußen (1911-2005), Princess Marianne von Preußen (1913-1983), and her husband Prince Wilhelm von Hessen (1905-1942) on the 30th of November 1936

 

My Internet search continued. I found an undated group photo of Wilhelm von Hessen including him and his four siblings. (Figure 11) He is pictured fourth from the left but once more the evidence is not clear cut that he is the person in the group picture.

 

Figure 11. An undated picture of Wilhelm von Hessen with his four siblings in civilian clothes

 

Another headshot of Wilhelm von Hessen in his military uniform, also undated, was not conclusive. (Figure 12)

 

Figure 12. An undated photo of Wilhelm von Hessen in his military uniform

 

Finally, I had the good fortune to find in Geni a high-quality picture of Marianne von Preußen and Wilhelm von Hessen, possibly from around the time they were married on the 30th of January 1933 (Figure 13); at the time, Marianne would have been 20 years old and Wilhelm 28 years, the approximate ages they appear to be in the picture. A simple visual comparison of this image of Wilhelm to the person in the group photo left me fairly convinced they were the same person. I asked several friends who I know from previous experience are adept with visual comparisons, and they agree. Unfortunately, the reader who originally sent me the group photo disagrees.

 

Figure 13. Wilhelm von Hessen and his wife Princess Marianne von Preußen possibly around the time they were married on the 30th of January 1933

 

While I have no expertise in this area, both readers agree the man in the 1936/37 group picture is wearing a Schutzstaffel or SS uniform. However, the informant who sent the group photo claims that Wilhelm von Hessen did not switch from the Wehrmacht to the SS (i.e., German soldiers could not be members of both units simultaneously) until right before the Nazi invasion of France on the 10th of May 1940. Unfortunately, I can’t independently verify when von Hessen joined the SS. On his 30th of November 1936 photo, Wilhelm von Hessen is supposedly wearing a Wehrmacht uniform. (Figure 14) Based on this apparent discrepancy, I theorize the group photo was taken in late 1936 or in 1937 after the November 1936 marriage, by which time Wilhelm von Hessen was then a member of the SS.

 

Figure 14. November 1936 photo of Wilhelm von Hessen and Princess Marianne von Preußen at Princess Luise’s marriage in his Wehrmacht military uniform

 

Having convinced myself that the amateur historian is correct that the person in the 1936/37 group picture is not Reinhard Heydrich but Wilhelm von Hessen, as a lark I decided to see whether I could find an artificial intelligence (AI) application which could strengthen my case. I found an AI-powered tool which allows me to do precisely what I was looking to do, namely, compare two faces to measure similarity. It is called “FaceShape” and below is the link to this tool: 

https://www.faceshape.com/face-compare

I’m admittedly not adept at using new technology, so the attraction of FaceShape is that it’s supremely easy to use.

Juxtaposing the images of Reinhard Heydrich and that of the person in the group photo originally sent by the first reader yields a low probability of only 27.04% that they are the same individual. (Figure 15) By comparison when I compare the known image of Wilhelm von Hessen from his sister-in-law Princess Luise’s 1936 marriage to the person in the group photo, FaceShape claims a 100% probability they are the same person. (Figure 16) Another comparison that yielded a 100% match was Wilhelm von Hessen’s photo from his sister-in-law’s marriage to one of him in his military uniform. (Figure 17) Readers can see the results of the various other images I contrasted. (Figures 18-20)

 

Figure 15. FaceShape comparison of Reinhard Heydrich and his doppelgänger showing only a 27.04% likelihood they are the same person

 

Figure 16. FaceShape comparison of a known photo of Wilhelm von Hessen to the person in the group picture indicating a 100% probability they are the same person

 

 

Figure 17. Another FaceShape comparison of two known images of Wilhelm von Hessen suggesting a 100% probability they are the same person

 

Figure 18. FaceShape contrast of two known images of Wilhelm von Hessen indicating a 92.27% probability they are the same person

 

Figure 19. FaceShape contrast of two known images of Wilhelm von Hessen indicating an 83.21% probability they are the same person

 

Figure 20. FaceShape contrast of two known images of Wilhelm von Hessen indicating a 63.71% probability they are the same person

 

Comparing the photos I found of Wilhelm von Hessen to one another, then to the person in the group picture from 1936/37, sets my mind at ease that it is indeed Wilhelm von Hessen. However, I want to be very clear with readers that FaceShape is not perfect. Based on some of the images I juxtaposed, the application obviously does not work well with poor resolution images; where the person in question is partially blocked by another individual; where siblings are involved that resemble one another or are of similar age; when known pictures of the same person are from earlier in life vs. later in life; or in distinguishing gender.

In closing, I would merely say that I consider FaceShape or similar AI-powered applications to be one tool in an arsenal that genealogists can use to further one’s research and possibly resolve thorny identification questions. It clearly requires human interpretation after the tool is applied to consider the question of whether the results are logical and make sense. While I may not have convinced the original reader that Reinhard Heydrich is not in the group photo taken at Castle Kamenz, I have proven to my satisfaction the person in question is Wilhelm von Hessen, not Heydrich.