Systems and methods for digital ink generation and editing

A computer-implemented method, comprising:

• training a handwriting model, wherein the training comprises:
• for each given training example of a plurality of training examples, the given training example including an original handwriting sample representing a sequence of characters and an original label indicating the sequence of characters represented in the original handwriting sample:
• processing, using the handwriting model, the original handwriting sample to generate a style vector having a plurality of elements;
• modifying, using one or more processors of a processing system, a random selection of one or more of the plurality of elements of the style vector to generate a masked style vector;
• generating, using the handwriting model, a predicted handwriting sample based on at least the masked style vector and the original label;
• and
• comparing, using the one or more processors, the predicted handwriting sample to the original handwriting sample to generate a loss value for the given training example;
• and
• modifying, using the one or more processors, one or more parameters of the handwriting model based at least in part on the loss values generated for each of the plurality of training examples.