Free web application ``Calligrapher.ai'' review that can generate ``handwritten characters'' with the power of AI
There should be many people who have experienced wanting to use 'English handwriting' while creating illustrations and slide materials. If you use the free web application ' Calligrapher.ai ', it will convert the input text into handwritten characters quickly, so I actually tried using it.
Calligrapher.ai: Realistic computer-generated handwriting
When you access Calligrapher.ai from the above link, the following screen will be displayed.
To generate handwritten characters, enter characters in the input field at the bottom of the screen and click 'Write!' At the bottom right or press the Enter key. This time I entered 'GIGAZINE'.
The generated handwritten characters look like this.
Even if you enter the same sentence, different handwritten characters are generated each time you click 'Write!'. Both of the following two images are handwritten 'GIGAZINE'.
You can adjust the readability by moving the 'Legibility' slider at the top of the screen. Readability minimum is below.
And the maximum readability is below.
You can also adjust the thickness of the handwritten characters by moving the 'Stroke Width' slider. The minimum line thickness is below.
The maximum line thickness is below.
In the 'Style' pull-down menu, you can select styles 1 to 9 (the habit of writing characters). The default '-' seems to pick the style randomly. Even if the same character is handwritten in the same style, the result will change each time it is executed, so it is recommended to generate it like a gacha until the handwritten character you like is generated.
You can download the generated result as an SVG file by clicking the download button on the upper left.
After that, you can process the downloaded SVG file and use it for your work.
The source code of the handwritten character generation algorithm used for Calligrapher.ai can be confirmed at the following link.
GitHub - sjvasquez/handwriting-synthesis: Handwriting Synthesis with RNNs ✏️