Matching digital fonts utilizing a residual neural network

A non-transitory computer-readable medium storing executable instructions, which when executed by at least one processor, cause the at least one processor to perform operations comprising:

• determining one or more textual characters within a target digital image;
• determining one or more predicted fonts for the one or more textual characters utilizing a font recognition machine learning model to extract features of the one or more textual characters from the target digital image, the font recognition machine learning model comprising parameters learned from synthetic text data including sample textual images generated with a multi-attribute probabilistic model by selecting, for a sample textual image of the sample textual images, two or more text attributes from a plurality of attribute categories corresponding to a distribution of text attributes;
• and
• providing, via a user interface on a client device, the one or more predicted fonts for the one or more textual characters.

A system comprising:

• one or more memory devices a font recognition machine learning model;
• and
• one or more processors configured to cause the system to:
• determine one or more textual characters within a target digital image;
• determine one or more predicted fonts for the one or more textual characters utilizing the font recognition machine learning model to extract features of the one or more textual characters from the target digital image, the font recognition machine learning model comprising parameters learned from synthetic text data including sample textual images generated with a multi-attribute probabilistic model by selecting, for a sample textual image of the sample textual images, two or more text attributes from a plurality of attribute categories corresponding to a distribution of text attributes;
• and
• present, via a user interface on a client device, the one or more predicted fonts for the one or more textual characters.