Restoring Faces in Your Photos Using GFPGAN- Machine Learning for Genealogy

Original Photo 

Restored Photo

As some of you know already, I am an IT major at Marymount University. Recently I have developed a persistent interest in machine learning and artificial intelligence, so you can imagine my excitement when I discovered a tool that incorporates these exciting and constantly developing fields with my other interest- genealogy.

The tool is called GFPGAN and its stated purpose is as a Practical Algorithm for Real-world Face Restoration.

Now before I begin nerding out on all the cool features and advanced ways you can use this tool, I want to drop a couple links that will make using the tool infinitely easier for those of you who -well- just want to use it. These online demos will let you try out the tools without going through a bunch of complicated steps.

Videos Honoring Timothy Wesselowski: His Life and Funeral Service

The Videos

Documentary on My Grandpa's Life

My Grandpa's Funeral Service

My Grandfather's Death

On the 2nd of January, 2021 my Grandpa Tim Wesselowski passed away at the age of 80 years old. He had contracted COVID-19 in the month before and died as a result of the accompanying pneumonia. Unlike many who met the same fate in the last year, he did not die alone. My Uncle Nathan was by his side and my Mom was on speaker phone. My Uncle had spent the prior day with him doing one of his favorite things: watching football with family.

A Tool to Change the Way You Conduct Collateral Research (Census2Ged v2.1)

Table of Contents

1. Links 

2. Introduction

3. Wait, What is Collateral Research?

4. So, How Does Census2Ged Help With Collateral Research?

5. Installing Census2Ged

6. Prerequisites
7. Using Census2Ged
     a. Gedcom Name
     b. Country
     c. Census Year
     d. Sourcing
     e. Source List Name
     f. Everything Else in Sourcing
     g. United States Tags
     h. Swedish Household Examinations
     I. Swedish Household Examination Tags
8. Cleaning Up the Final Gedcoms

Links

1. Like Census2Ged? Consider sending a couple dollars to my PayPal to support its continued development: https://www.paypal.me/ReneeSchmidt

2. Census2Ged on Github: https://github.com/xXReneeXx/Census2Ged

3. Download Census2Ged: https://github.com/xXReneeXx/Census2Ged/releases/tag/v2.2

4. Join the Census2Ged user group on facebook: https://www.facebook.com/groups/1386416641488942/

Introduction

Eleven months ago I published version 1.0 of my software, Census2Ged. It was buggy, had no customizability, and had an appearance reminiscent of the dark days where graphic design and software had an oppositional relationship.

Census2Ged 1.0

I am happy to announce that those days of visual and operational vulgarity are now over, thanks to the advent of the latest version, Census2Ged v2.1.

Census2Ged 2.1

The functional differences between v2.1 and all previous ones are massive, and the program will change the ease of conducting collateral research for all who use it.

Using FTDNA's API: Commands and Explanations

After hearing from all of you, I am now fairly certain that anyone is, in fact, allowed to access FTDNA's family finder API. And, as a bonus, I learned some new things about how APIs function in websites as well (special thanks for that goes to the anonymous commenter on the previous post). The question that remains is whether or not programmatically accessing it is against FTDNA's terms of use, and if anyone knows anything about that I'd be very happy to hear it, as I can think of a few things I want to try doing with the API and my best friend, python.

Anyways, this new post will go over the ways you can use the API and the different requests you can send to it. It's not all-inclusive and only includes the things I thought were important/most useful.

Please note, this post will include a lot of technical words without a lot of detailed explanation. If you just want to try it out for yourself simply click one of the links in the post to see the match data after signing into ftdna in your web browser.

How to Send GET Requests (In General)

There are two ways you can send Requests. The first is to just put the URL in your browser and the 2nd is to use Postman. For the most part, I use because making requests in it is faster and its easier to change parameters, however, I will use my browser in this tutorial because that will make it easier to collapse things/ limit the amount of censoring I have to do.

Match Lists

Basic Match List

The JSON Data for the Basic Match List

To return the information for your first 9 matches in JSON format (the same number as you would normally see on the screen) Use the following url:

https://www.familytreedna.com/my/family-finder-api/matches?

Full Match List

You'll notice that the first line of JSON data returned by the last command has the total count of matches you have. You can use this number or any number larger than it in the url to print out that number of matches. In the url set page=1 (to start on your first page of matches) and set pageSize=numberofmatches. Here is an example url:

https://www.familytreedna.com/my/family-finder-api/matches?page=1&pageSize=4000

Using FTDNA's API

Edit 1 5/11/2018: Note that this does not mean anything bad for you at all and you can continue using ftdna as normal, it's a fantastic site. Also, there is a possibility that the information I was sent regarding who was supposed to be able to use the API could have been incorrect, see the comments section for an interesting discussion on that. I think that the big takeaway from this post should be if you want to use the API for your company it is worth checking in with ftdna to ensure you have permission to do so.

Edit 2 5/11/2018: I have been looking into it more, and I believe that this is not in fact an issue at all which I am very happy about. I am still confused about why ftdna told me otherwise, and it may be that you are technically still not supposed to use it, but I'm really not sure at this time. Special thanks to the anonymous commenter for clearing things up!

Note: I actually wrote this all the way back on March 12th. I had notified FTDNA of the fact that their API was accessible by the public even before that. I am publishing this now because I feel it is my right/duty. They made the decision not to fix it or notify the public that they can use the API, so I think that somebody should. I'm hoping that this post will perhaps raise awareness about this and make FTDNA make a definitive choice about whether they want their API to be publicly available (which I personally think would be fantastic).

Purpose

This document will serve to explain the process through which I found a security vulnerability in the FTDNA web API, as well as the process I went through in reporting said vulnerability and what I would have done differently if given the chance.

Definitions

Some company names and terms used in this document will undoubtedly be unfamiliar to those reading it. This section will serve to clarify said terms.

How OneNote Can Help You Transcribe Documents

OneNote is my program of choice for transcribing newspapers. It's OCR tools, instant copy paste photo insertion, and flexible formatting functionalities make it a breeze to transcribe documents in no time at all. This post will take you through my transcription workflow, allowing you to just as easily transcribe your source material in a matter of minutes.

What is OCR and Why Should You Use It?

OCR stands for Optical Character Recognition. A character is a single letter within a word. OCR programs will 'look' at an image and identify the characters inside, allowing you to copy and paste text from an image. Using OCR programs will make transcribing go much quicker since the computer does most of the work, allowing you to spend more of your time researching and making discoveries.

1. Insert Your Source Image into One Note

There are lots of ways you can do this, but I will just show an example of one. Much of the time I am not transcribing a full image, but just a portion of one. Thus I usually only want to import a portion of an image into One Note. Windows operating system has a nice little tool called 'snipping tool' which will capture just a section of your screen for you, which is exactly what we need.

For my main image, I will use page 40 from the 1931 Bethany Daisy, my largest currently ongoing transcription project. Here is the image if you wish to follow along:

Page 40 of the Bethany Daisy

\

Part 1:The Complete Guide to Deciphering Kurrentschrift (For Non-German Speakers)

Oftentimes the most difficult task for anyone researching their German ancestors is navigating Kurrentschrift.

Kurrentschrift is an old German script which fell out of use after 1941. Although the script is quite beautiful, it can be very difficult to read when first starting out (and for some time after as well). This is the first post in a series which will (hopefully) teach you everything you need to know to start deciphering these texts. It's important to note that these will only give you a jump start. You will be able to (hopefully) read most of your records after working through the steps I outline, however it will take a lot of time at first to read each one. Speed comes with a combination of time and recognition, both of which mean you have to practice. Keep at it though, with each record you transcribe your time spent will decrease exponentially. Personally, transcribing records written in Kurrentschrift (as long as the handwriting isn't smudged or cramped) now takes only a few more minutes than transcribing English records.

Now with that in mind, let's get started!

Using Schrift Generator

Hopefully, the future sections of this series will help you to read your documents in Kurrentschrift relatively well without the use of technology. However, no one can deny the value of having a bit of help here and there. That's where the Schrift Generator comes in handy. The Schrift Generator  allows you to type out anything you want and display how it would look when written in a variety of Fraktur, Sütterlin, and Kurrent fonts. 

The User Interface

The page is in German, but it's relatively easy to use. If you scroll down, you should see an entry field that looks something like this:

The text entry field in Schrift Generator

This is where you type the words you wish to display in German Script.

Visualizing Shared Events in Your Tree

When I conduct collateral research, I like to add neighbors, friends, and possible relatives directly to my family file, linking them together using shared events. This increases the chance that I will accidentally happen across a family member, while also giving me an idea of their stance within the community. I decided to figure out how to display shared events in a way that I could see all the connections I'd established for a person with a single click. I found a completely free solution and will explain in the rest of this post what that solution was. These network graphs were the result of my efforts.

Network graph of all the people in my file based on their shared events

Filtered view within the network graph of just one person in my file and their connections

A Progress Update on Census2Ged

I am not quite ready for a release yet, but the next release of my free program, census2ged, will have insane amounts of new functionality added. Since I'm so excited that it's starting to look like a real program, I thought I'd share some of what's to come.

If you'd like to read my previous posts about the program (which I programmed in python and have made available to anyone for free) you can check out the following links:

https://famgenealogy.blogspot.com/2017/09/census2ged-my-new-program.html
My first post about it. It goes over the basics and the functionality included in the first release.

https://famgenealogy.blogspot.com/2017/10/census2ged-20.html
My second post which details the new things I added in the second release.

Anyways, let's get to it! And, if you have any suggestions of things I should add before the actual release just let me know in the comments!

New and Improved GUI

That's right, it no longer has only two input fields!

Okay, so it still looks a little like the neglected lovechild of 90's design and Windows Vista, but we can ignore that for now. 

A Comprehensive Guide to the Old Fulton New York Postcards Website

On December 25th the Old Fulton New York Postcards website (which, by the way, is not a website for postcards) received a major update that added a whopping 700,000 more old newspaper articles from the USA and Canada. The site, however, has an aesthetic that makes it slightly more difficult to start using than the average site. This post will help you learn to use this incredible resource to its full capacity.

Old Fulton New York Postcards homepage

3 Ways to Use Archive.org for Genealogy

Archive.org is one of the most valuable free resources for genealogy you have at your disposal. This post will teach you how to use it for your own research.

Internet Archive Home Page

Census2Ged: My New Program

I was thinking about how I could get past some brick walls in my research using neighbors to my relatives. I ended up transcribing several census pages into Genscriber and then entering everyone on them into my genealogy program as neighbors on a shared census event. However, this was tedious and took way more time than I could reasonably afford it. So I started to think about how I could use my programming knowledge to solve the problem. I wanted to be able to create trees organically using censuses while preserving nuclear family relationships. 3 days ago I started working on Census2Ged. It only works with the 1900 census as of now, but I hope to extend its functionality to other censuses in the future. This post will show you how to use it and what it produces. And, of course, you can download it for free at this link: https://sourceforge.net/projects/census2ged/

Features:

These are the features currently supported in Census2Ged:

• Works with the 1900 US census (and in the next update it will work with all censuses from 1850-1940) 
• Conforms to gedcom 5.5 Standards
• Preserves relationships between heads of the family, wives, and their children
• Records first and last names in correct fields
• Records Race and sex
• Records birth month and birth year
• Records approximate year married
• Records how many children they have living and how many are dead
• Records year of immigration and naturalization information
• Records Occupation
• Records Literacy information
• Records Property information

It does not yet support linking stepchildren, parents of heads, or other non-standard relationships.

A completed individual file looks like this:

Writing Master Source Names in Legacy Family Tree

I recently realized that I really needed to start getting my Master sources straight. The Source Writer is the main reason why I love Legacy Family tree. It makes writing high quality sources super easy and fast. However, family files will very quickly accumulate a large number of master sources in a very short amount of time. That's why organizing them and coming up with naming conventions is so important, which is what I will show you how to do today.

This is what my sources look like currently. As you can see they are not really organized and the naming conventions vary from source to source. You'll notice that none of my sources are person-specific. That is just personal preference; I use each master source dozens of times for multiple people to save time.

The first thing I will do is come up with naming conventions. The exact format I prefer is: "Date RecordType State/Country County Website". After renaming all of my master sources for the 1910 census, this is what it looks like:

As you can see I have used star signs, dash marks, and parenthesis to separate the record type, location, and source website from each other. I think this increases readability and makes it easier to immediately spot the one you are looking for.

This format organizes the master sources in a very specific way:

1. It sorts your sources by date
2. It then organizes further by the type of record
3. Next it sorts by either the state or country it is in. You want the state first so that all the Master Sources for that state and date stay by each other. Otherwise it will organize alphabetically by the county
4. It organizes alphabetically based on the original website the master source is for.

Let me know in the comments how you organize your sources, or tell me how this method works out for you!

The future of Genetic Genealogy

I recently came across this article by ancestry.com. A quick summary: Ancestry was able to reconstruct a partial genome for a man named David Speegle and his two wives using the shared segments of his descendants. The article finishes with "AncestryDNA (will probably not) help reconstruct the DNA of your great-great-great-grandfather". This is probably true since Ancestry currently does not incorporate a chromosome browser into their DNA package. It does, however, bring up the question: would it be possible to reconstruct the DNA of our ancestors on a larger scale?

How to track Twins in Legacy Family Tree

As many of you may know, the tendency of mothers to have fraternal twins is a genetic one. Fraternal twins run in nearly all lines of my family tree. In fact, I am one myself! I recently decided to start tracking these twins using Legacy Family Tree and decided to share the steps to do this.

There are 2 main options for tracking twins in Legacy Family Tree. The first is by marking their 'child status' as a twin. I will show the steps for this using my great grandmother, Lillian Penas, and her twin sister, Helen Penas.

First, navigate to the "family view" of the person you wish to mark as a twin.

The family view for Lillian Penas

How to use github to host your genealogy website for free (for complete beginners)

I was looking for ways that I could host my small, work in progress, genealogy website for free. All the options seemed to have negatives that outweighed the positives, or had limited storage amounts and no guarantee of long term hosting. Then I remembered Github pages. This will be a full tutorial on how to do what I did for complete beginners. It will assume you have a website generated by your genealogy program. (I recommend GRAMPS as a good free program.)

The end result will be something like this (my website):

https://reneeschmidt.github.io/Genealogy/individuals.html

What is Github?

Github is a code sharing website. It is mainly for open-source projects, which means projects where the code is free for viewing by anyone who wants to.  This is the link to the github homepage: https://github.com/ .

Getting Your Github Account

Simply go to github and create a free account, Its that easy!

Installing Everything You Need

Downloading GIT

GIT, unfortunately, does not work with just the website. You will need to install an application on your computer in order to get it to work.

First go to the following link: https://git-scm.com/downloads

Click on the operating system that you are working on. In my case its windows so I clicked the “Windows” button.

What I've learned from scanning in excessive amounts of photos

To say I have a lot of photos would be an understatement. I bought myself a 1 terabyte hard drive to keep all of them on, and I've already used up a quarter of the space. About 30,000 of these photos are of me and my sisters (my dad was a bit excessive with the photos when we were little), and about 700 are photos I've scanned in. The first hundred or so I scanned in using the scanner on our printer.

Our printer, as I quickly discovered, is meant for printing- not scanning. It moved at a snail pace and scanned in at a max of 600 dpi, which is fine for most pictures but with certain exceptionally tiny pictures a larger dpi is necessary. It also only scanned pictures in as jpegs. I don't like jpegs at all. They are too unpredictable for scans in my opinion and since so much data is compressed in them and detail lost they reduce the maximum possible quality and detail I can instill in them with photoshop. I much prefer png or tiff. Another problem with my printer scanner was it only scanned in 1 picture at a time and the automatic cropping was poorly done. Sometimes I'd have to put construction paper under my photos so it didn't crop off parts of the picture and crop manually later on.

Luckily for me my dad mentioned an old flatbed scanner we had in our basement. As soon as I started using it I knew I could never go back. Its an Epson Perfection 3490 photo and its perfect. Of course when I first started using it I decided I would scan every photo in at the max dpi. My tip to you is: don't do this. Every photo was half a gb and it was a pain to resize them all. I usually go with 700 for small Polaroids and 400-500 dpi for the more modern and larger Kodak film. But, once I got that figured out it was amazing. I can scan in and label 300 photos in a few hours because it allows you to scan in multiple photos at once. And its speedy. If you don't have a good scanner like this it is well worth the investment.

And now I'll finish with some of my favorite photos I've scanned of the Wesselowski / Byers side of the family. (most of these have been photoshopped for quality)

Helen Byers Basketball Captain 1926 as a senior in high school.