Methods, systems, articles of manufacture, and apparatus for image recognition based on visual and textual information

At least one non-transitory machine-readable storage medium comprising machine-readable instructions to cause at least one programmable circuit to at least:

• classify object facings detected in an image to predict object candidates corresponding to a first object depicted in the image, ones of the object candidates to include a respective image-based score;
• group first words extracted from the image, the first words corresponding to a first one of the object facings;
• generate text-based scores for the object candidates by comparing respective object characteristics of the object candidates to the first words, ones of the text-based scores corresponding to respective ones of the object candidates;
• determine final scores for the object candidates by selecting, for the ones of the object candidates, (a) a respective image-based score or (b) a combination of the respective image-based score and a respective text-based score;
• and
• arrange the object candidates based on the final scores.

A method comprising:

• classifying, by at least one programmable circuit programmed by at least one instruction, object packaging detected in a shelf image to determine object candidates corresponding to a first object, ones of the object candidates to include a respective first metric, the first metric based on visual features of the object packaging;
• associating, by one or more of the at least one programmable circuit, first words extracted from the shelf image to a first one of the object packaging;
• generating, by one or more of the at least one programmable circuit, second metrics for the object candidates by comparing respective object characteristics of the object candidates to the first words, the second metrics based on text, ones of the second metrics corresponding to respective ones of the object candidates;
• determining, by one or more of the at least one programmable circuit, third metrics for the object candidates by selecting, for the ones of the object candidates, (a) a respective first metric or (b) a combination of the respective first metric and a respective second metric;
• and
• sorting, by one or more of the at least one programmable circuit, the object candidates based on the third metrics.