Segmentation of an input by cut point classification

A system comprising:

• one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
• obtaining, for a sequence of strokes that represent a handwritten input, cut point data indicating one or more particular candidate cut points that are identified within the sequence of strokes;
• obtaining, for the one or more of the particular candidate cut points, feature data indicating one or more features of the particular candidate cut point;
• for each of the one or more particular candidate cut points, providing the feature data to a classifier that is trained to predict, based on one or more features of a candidate cut point, a likelihood of the candidate cut point being a correct cut point;
• for each of the one or more particular candidate cut points, receiving, from the classifier, data indicating the likelihood that the particular candidate cut point is a correct cut point;
• selecting a set of one or more of the particular candidate cut points whose respective likelihoods satisfy a threshold;
• and
• using the set of candidate cut points to segment the sequence of strokes.

A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

• obtaining, for a sequence of strokes that represent a handwritten input, cut point data indicating one or more particular candidate cut points that are identified within the sequence of strokes;
• obtaining, for the one or more of the particular candidate cut points, feature data indicating one or more features of the particular candidate cut point;
• for each of the one or more particular candidate cut points, providing the feature data to a classifier that is trained to predict, based on one or more features of a candidate cut point, a likelihood of the candidate cut point being a correct cut point;
• for each of the one or more particular candidate cut points, receiving, from the classifier, data indicating the likelihood that the particular candidate cut point is a correct cut point;
• selecting a set of one or more of the particular candidate cut points whose respective likelihoods satisfy a threshold;
• and
• using the set of candidate cut points to segment the sequence of strokes.