Multimodal input contextual font recommendations

A non-transitory computer-readable storage medium including instructions stored thereon which, when executed by at least one processor, cause the at least one processor to:

• receive a modal input including at least one of a text input or an image input;
• extract a set of intents from the modal input;
• generate, by an intent embedding generator including a machine learning model, an intent embedding from the set of intents;
• compare the intent embedding to a plurality of candidate font embeddings to obtain one or more candidate fonts based on a similarity of the intent embedding to the plurality of candidate font embeddings, wherein each candidate font embedding of the plurality of candidate font embeddings represents an intent of a respective font, wherein each candidate font embedding of the plurality of candidate font embeddings is based on text representing the intent of the respective font, and wherein candidate font embeddings of the plurality of candidate font embeddings that are closer in distance to each other in an embedding space represent fonts having more similar intents than candidate font embeddings of the plurality of candidate font embeddings that are farther in distance from each other in the embedding space;
• and
• identify a recommended font based on the similarity of the intent embedding to a selected candidate font embedding of the plurality of candidate font embeddings.

A system comprising:

• a set of one or more processors;
• and
• a memory including instructions which, when executed by the set of one or more processors, cause the system to:
• receive a modal input including at least one of a text input or an image input;
• extract a set of intents from the modal input;
• generate, by an intent embedding generator including a machine learning model, an intent embedding from the set of intents;
• compare the intent embedding to a plurality of candidate font embeddings to obtain one or more candidate fonts based on a similarity of the intent embedding to the plurality of candidate font embeddings, wherein each candidate font embedding of the plurality of candidate font embeddings represents an intent of a respective font, wherein each candidate font embedding of the plurality of candidate font embeddings is based on text representing the intent of the respective font, and wherein candidate font embeddings of the plurality of candidate font embeddings that are closer in distance to each other in an embedding space represent fonts having more similar intents than candidate font embeddings of the plurality of candidate font embeddings that are farther in distance from each other in the embedding space;
• and
• identify a recommended font based on the similarity of the intent embedding to a selected candidate font embedding of the plurality of candidate font embeddings.